free trial issue                                  subscribe                                             back issues

Athena Review Vol. 5, no. 1


Records of Life: Fossils as Original Sources


Middle and Late Permian Therapsids 

        In the Permian period, all major land masses were joined in the mega-continent Pangea, which formed from the juncture of two supercontinents, Laurasia in the north and Gondwanaland in the south. These had gradually assembled during the preceding Devonian, Mississippian, and Pennsylvanian periods.  Pangea lasted through the Permian and into much of the subequent Triassic period, before the rocky plates representing the continents began to separate.

            A key indicator of the connection of land masses in the Permian is the distribution of the same species in continental areas now separated by oceans.  A good example of this is the conifer Glossopteris, an important Permian forest element preserved in the fossil record. Glossopteris was a woody, seed-bearing tree, growing as high as 30 meters, and interpreted to have thrived in very wet soil conditions, such as around swamps and along river banks. Fossils of Glossopteris have been found in all of the southern continents (fig.1), with over 70 species identified in India alone, and other species known from Africa, Madagascar, Australia, Antarctica, and South America.  Appearing by the Early Permian period (298-280 mya), members of the Glossopterid family became the dominant elements of the southern flora through the rest of the Permian. They then disappeared during the Permian-Triassic mass extinction of 252-250 mya.

Fig.1: Distribution of Glossopteris (green area) in southern Pangea.

         The zone of Glossopteris distribution (fig.1) across several, now detached, landmasses led Austrian geologist Eduard Suess (1885-1909) to infer that these areas had once been connected by a land bridge, with the continents remaining in the same position.  As expressed in his major work, The Face of the Earth (1885-1909), Suess also believed that the rise and fall of sea levels were mappable across the earth through geological time, and that the periods of ocean transgression and regression could be correlated from one continent to another. He named the interconnected southern land mass Gondwanaland, after the district in India where fossils of the plant Glossopteris were abundantly found. The ancient sea that he postulated to be north of Gondwanaland he named the Tethys Sea.

        Shortly thereafter,  the German meterologist and polar researcher Alfred Wegener (1880-1930) more accurately interpreted that the land masses themselves had moved together by shifting continental plates. Wegener's theory of Continental Drift (1912, 1929), hypothesizing that the continents were slowly moving around the Earth, was not widely accepted for several decades. Among its most vocal opponents was the American paleontologist George Gaylord Simpson (1943).  Wegener's theory began to gain wide acceptance in the early 1950s, when paleomagnetic samples taken from India showed that the country had previously been in the Southern hemisphere, as Wegener had predicted. His theory is now fully confirmed and forms the starting point for current models of plate tectonics. Suess's names Gondwanaland and the Tethys Sea have also both been retained. The name Pangea, from pan ("entire" [Gk]) and Gaia (the name of the Greek earth goddess), thus "the whole earth", was coined at a 1927 conference of geologists held in Tulsa, Oklahoma to discuss Wegener's theory of continental drift (Willem et al. 1928). 

            Pangea is the setting for the first appearance in the fossil record of mammal-like reptiles or theriodonts, the ancestors of mammals, during the Middle and Late Permian (268-251 mya). The clearest exposure of this lies in a portion of the southern region of Gondwanaland, centering in South Africa.

.

South Africa: Karoo Basin  

        The best continuous fossil record from the Middle Permian through Early Triassic periods has been found in the Karoo Basin of South Africa. This sedimentary basin is bounded on the south by the Cape Fold Mountains, which were uplifted in the Early to Middle Permian period by tectonic plates converging on those of South Africa (Rubridge 1995). 

        The geological formations in the Karoo Basin include a succession of five groups (Dwyka, Ecca, Beaufort, Stormbert, and Drakensberg), dating from the Late Carboniferous through Jurassic periods (fig.2). The third of these, the Beaufort Group, is the lowest terrestial formation, overlying shallow marine shales of the Ecca Group. Beaufort Group rocks consists of shales and sandstones that represent riverine deposits of an ancient sedimentary basin. Its importance for paleontology stems from its having provided  a comprehensive series of often well-preserved fossil zones, including a continuous record of flora and fauna dating from the Middle Permian through Middle Triassic periods (270-220 mya) (Rubridge 1995). 

Fig.2: Geological formations in the Karoo Basin of South Africa.

         Fossil evidence of Therapsids ("beast faces"), the ancestors of mammals, first appear in the Middle Permian deposits of the Beaufort Group in South Africa.  Three suborders of Therapsids will be briefly traced below in their appearance in the Beaufort faunal zones: the Anomodontia ("undefined teeth"), the Dinocephalians ("terrible heads"),  and the Theriodonta ("beast teeth").  The theriodonts became one of the two synapsid survivors of the great Permian–Triassic extinction event, the other being the dicynodonts, part of the Anomodontia. Theriodonts split into two groups, Therocephalians ("beast heads") who died out after the Early Triassic; and cynodonts ("dog teeth"), whose carnivorous forms became progressively smaller during the Triassic. By the Late Triassic the first mammals evolved from small, shrew-sized cynodonts called trithelodonts.

        The Beaufort group has been subdivided into eight faunal assemblage zones (A.Z.), each associated with one or more local geological formations, and each containing a number of associated taxa (fig.3). 1. Eodicynodon A.Z., Abrahamskrall Formation; 2. Tapinocephalus A.Z., Abrahamskraal Formation; 3. Pristerognathus A.Z., Koonap and Middleton Formations; 4. Tropidostoma A.Z. Middleton Formation; 5.  Cistecephalus A.Z., Middleton and Balfour Formations;  6. Dicynodon A.Z., Balfour Formation;  7. Lystrosaurus A.Z, Balfour and Katberg Formations;  8. Cynognathus A.Z., Burgersdorp Formation (Rubridge 1995).

       1. Eodicynodon Assemblage Zone. This dates from the Middle Permian, Wordian stage (268.8-265.9 mya), and occurs at the southwestern edge of the Abrahamskraal Formation, the lowest zone of continental rocks in the Karoo Basin. This is composed of thick layers of fluvial-deposited shale or mudstones totalling 1400-1800 m in depth, with a higher percentage of sandstone, limestone, and chert than in the overlying Teekloof Formation (Keyser and Smith 1978). 

        The Abrahamskraal layers,heavily folded to the south, directly overlie marine-deposited shale of the Ecca Group's Waterford Formation. The intersection of the Waterford and Abrahamskraal Formations represents the paleoshoreline at about 270-269 mya (Rubridge 2000; Modesto et al. 2001).  

Fig.3: Stratigraphy of  the Beaufort Group Faunal Zones. Skulls used as examples: 1. Eodicynodon. 2. Robertia. 3. Pristerognathus. 4. Lycaenops 5. Cistecephalus. 6. Dicynodon.  7. Lystrosaurus.  8.  Cynognathus.       

            The zone is named for the small herbivore Eodicynodon ("dawn double dog-tooth"), the earliest known genus of dicynodont, first found as a skull at Farm Zwartskraal near Prince Albert (Barry 1974). Subsquent recovery of postcranial elements revealed that the limbs of Eodicynodon show some relatively primitive (i.e., reptilian)  traits, in that they tended to sprawl in standing posture. Eodicynodon was about 15 cm tall at the shoulders, and 45 cm in total body length.  It had strong forelimbs and broad paws (mani) with claws, indicating it dug frequently for food.  The hind legs were longer and more gracile, which is a typical derived feature for dicynodonts (Rubridge, King, and Hanson 1994). 

            Eodicynodon is relatively similar to the herbivore Robertia, also commonly found in this zone, although the two genera can be distinguished by the differing forms of the humerus or upper arm bone.  Other Therapsids found in this zone include Patranomodon Tapinocaninus, and Australosyodon. The latter two genera are part of the Anteosauridae family, considered ancestral to the Dinocephalia from the Late Permian.  Australosyodon is considered the most primitive Dinocephalian yet found in Gondwanaland (Rubridge 1994). Related genera of Anteosauridae include the Russian taxa Paranteosaurus, Sinophoneus and Titanophoneus.
            2. Tapinocephalus Assemblage Zone  (Middle Permian, Capitanien stage, 265.8 -261.2 mya). This, the thickest biozone in the Beaufort Group, was originally divided into three layers, a scheme that has subsequently been refined.  The lowest division contains many dinocephalians and therocephalians, with relatively few dicynodonts. The middle division has many dicynodonts and therocephalians, but relatively few pareisaurs and dinocephalians. The  topmost layer, with a scarcity of faunal elements, includes therocephalians and dicynodonts, but no dinocephalians (Boonstra 1969).
           Revision of the Tapiocephalus zone separated out this topmost layer, which now comprises the Pristerognathus A.Z. (Keyser and Smith 1978; Rubridge 1995). 
           Extensive deposits of t
his zone occur at the western side of the Karoo Basin, along the interface of the Ecca and Beaufort Groups, where the strata are often complicated by excessive folding of the shale layers, thus making it difficult to identify the sequence and chonology of fossil-bearing deposits. 
        In the lower part of the
Tapinocephalus Zone, three main families of dinocephalians are represented: the Anteosauridae ("earlier lizards"), also found in the underlying Eodicynodon Zone; the Titanosuchidae ("Titanic crocodiles") thought to have descended from the Anteosauridae; and the Tapinocephalidae ("humble heads"), including the type genus for the zone, Tapinocephalus (Rubridge 1995).

Dinocephalians:

Anteosauridae 

              The huge, semiaquatic carnivore Anteosaurus magnificus ("great earlier lizard"), the type species for the Anteosauridae, was named by Watson in 1921.  Up to 5 meters in length, and with a notably long skull (fig.4), it had prominent interlocking incisors, large canine teeth, and ten additional cheek teeth on each side (Boonstra 1954). Anteosaurus probably lived as a river predator, something like a modern crocodile. The giant Anteosaurus existed in South Africa at the same time that Titanophoneus and Doliosauriscus dominated riverine environments in the Isheevo region in western Russia. 

Fig.4: Skull of Anteosaurus magnificus, from the Mid Permian.

            Anteosaurus, defined by a relatively lage sample well-preserved skulls, shows distinctive traits in both its elongated, primitive skull form, and enlarged canine teeth which were related to its carnivorous feeding habits.  These traits were passed on to its later descendants such as the Titanosuchidae, many of which, however, changed their dietary habits to become omnivores or herbivores. In terms of their roles as carnivores, the family Anteosauridae were replaced in the Late Permian by large gorgonopsians (described below).

Titanosuchidae

            The second family of dinocephalians from this zone are the Titanosuchidae, considered direct descendants of Anteosauridae. Titanosuchidae include the genera Titanosuchus and Jonkeria, whose skulls show close similarities, retaining the primitive form of the anteosaurs, including a thickened frontal bone at the top, with a large pineal foramen.  The skeletal anatomy among the Titanosuchidae, however,  often differs significantly between genera, as presumbly did their eating habitsThe type species, Titanosuschus ferox ("fierce Titanic crocodile") is considered to have been a carnivore (King 1988). It measured over 2.5 meters in length, had a massive skull, and had curved, fang-like canines and sharp incisors, like those of Anteosaurus.

            A second genus from the same family, Jonkeria, is thought to have been a partial herbivore, sometimes eating plants (King 1988). Jonkeria ingens (fig.5), defined by Broom (1929), was 4-5 m long, almost as large as Anteosaurus, with long, stout limbs. It had a typically long, dinocephalian snout, with large incisors and canines. A total of six Jonkeria species have been named (Boonstra 1969).

Fig.5: Skeleton of Jonkeria ingens.

Tapinocephalidae

            The third family from this zone, the Tapinocephalidae, are also known in Russia, and probably had an even wider distribution in the Middle Permian, before becoming extinct at the end of the Capitanian stage. Primarily herbivores, their short, high skulls contrast significantly with the long, primitive skulls of the prevous two families (Anteosauridae and Titanosuchidae).  The Tapinocephalidae were among the largest animals of their age, weighting up to 3,000 - 4,500 pounds, with large, rounded midsections, as is typical for plant-eaters. Their teeth have chisel edges, and (also in contrast to the previous two families) lack the specialized canines of carnivores. Their relatively short, sturdy forelegs extended outwards, while the longer hind legs were placed directly under the hips (similar to that of the Late Permian  dicynodonts). Based on comparisons with some modern herbivores, they could stand bipedally and eat vegetation directly from lower tree branches.

            Tapinocephalus atherstone (fig.6),, the type species for this family, and the only known species of the genus, was a large herbivore up to three meters long, weighing an estimated 3,300-4.400 pounds. Their skulls are characterized by massive frontal bones in the skull roof, and short snouts.  

Fig.6: Skull of Tapinocephalus atherstonei, shown in overhead and lateral views (after Smith and Keyser 1995, fig.12a).

             Another large herbivore from the same family was Moschops capiensis (Broom 1911). Moschops ranged from 2.5 - 5.0 m in length, with a large, rounded midsection. They had a high, short skull, with short jaws and chisel-shaped teeth. There are two widely-accepted species, M. capensis and M. koupensis, both found within the same stratigraphic range of the Beaumont Formation.

                The pareisaurs are represented in this zone by Bradysaurus (figs.7,14), another large herbivore. Bradysaurus had heavily armored scales on its head and neck for protection against gorgonopsians and other predators. 

Fig.7: Skeleton of Bradysaurus (Berlin Mus. Nat. Hist.)

             3. Pristerognathus Assemblage Zone (late Middle Permian). This narrow zone, skirting the eastern edge of the much larger Tapinocephalus zone contains the first evidence of the carnivorous Gorgonpsida ("gorgon faces"), the most primitive group of theriodonts ("beast teeth"). The other two groups, Therocephalia ("beast heads") and Cynodontia ("dog teeth") are grouped as Eutheriodonts, "true beast teeth". The Eutheriodonts have larger skulls, with larger brains related to improved hearing and vision, and improved jaw muscles related to enlarged lower jaw bones (dentaries), a mammal-like trait which enabled more effective chewing. The early theriodonts were carnivorous; later, during the Triassic, several groups became herbivorous.

            The zone is named for Pristerognathus, a medium sized therocephalian carnivore discovered by Broom in 1904, growing up to 1.5 meters in body length. These otter-sized animals had long, narrow skulls with large canines, and probably hunted smaller therapsids and reptiles.  Three species include P. baini, P. polyodon, and P. platyrhinus. 

           At this point, another major change was occurring in the theriodonts, in the miniaturization of several smaller bones in the rear of the lower jaw (the reptilian quadrate, articular, and surangular bones),  and their transformation into the tiny ossicle bones of the mammalian middle ear (stapes, incus, and malleus). This process, completed by the Late Triassic,  led to the much improved hearing of mammals, as compared to reptiles who have only the stapes in the middle ear to amplify airborne sounds. Improved hearing helped make the theriodonts the most successful group of synapsids. The two diagnostic traits of the middle ear with three ossicles, and the related changes in the lower jaw joint, are the two most commonly used guides in the fossil record to distinguish early mammals from non-mammal synapsids as well as reptiles.

            4. Tropidostoma Assemblage Zone (Late Permian). This zone, dating from the Late Permian (Tatarian stage) is named for the early dicynodont, Tropidostoma ("keel mouth"), first reported by Seeley (1889) and Broom (1915) from Tafelberg in the Beaufort West District of South Africa. One species is known, T. micrtotrema, with various earlier names (Keyser and Smith 1978).  The Tropidostoma zone sees an increase in gorgonopsians as a primary predator.

            5. Cistecephalus Assemblage Zone  (Late Permian). This zone is named for the small dicynodont  Cistecephalus, one of the first known dicynodonts (Owens 1876, 1879)Cistecephalus was a small, burrowing animal up to 60 cm  long. It had a flattened and wedge-shaped skull, slightly convex as in present day burrowers, for the attachment of powerful neck and shoulder muscles.  Its strong forelimbs show structural similarities to those of modern burrowing mammals (King, 1990), such as the mole (Talpidae family).  The Cistecephalus Zone shows a peak in dicynodont diversity, including about 35 genera. By this time in the Late Permian, the larger dicynodonts have radiated to other parts of Gondwanaland (King 1990). This genus has also been identified in India (Bandyopadhyay 1988) and Zambia. A very similar genus, Kawingasaurus, is also known from the Kawinga Formation of Tanzania, considered equivalent to the Cistecephalus zone.  
       
In this zone,
the gorgonopsians show their largest presence, with the greatest number of their species identified.  

            6. Dicynodon Assemblage Zone  (Latest Permian, 257-252 mya). This extensive, ring-shaped zone  is named for Dicynodon ("Two Dog-teeth"), the type species of the dicynodonts, first described in the mid 19th century. It is a medium-sized, herbivorous therapsid about 1.2 meters long, with no teeth except for two prominent, tusk-like canines (fig.8). It probably cropped vegetation with a horny beak, much like a tortoise, while the tusks may have been used for digging up roots and tubers. The type species is Dicynodon lacerticeps, named by Owen (1845). Although over 160 species of Dicynodon have since been named from various Late Permian formations in Russia , China, India, and elsewhere, these have recently been reclassified. A recent study of the genus determined the only valid members of the genus to be D. lacerticeps from South Africa, and  D. huenei from Tanzania, with nine other species from other regions assigned to the infraorder Dicynodontia (Kammerer and Angielczyk 2009).

Fig. 8: Skull of Dicynodon.

            7. Lystrosaurus Assemblage Zone  (Early Triassic; 252-247 mya). While most of the Permian species described from the previous six faunal zones were eliminated by the massive extinction event at the end of the Permian, some dicynodonts and cynodonts from the Karoo Basin survived,  and their fossils are found in Early Triassic layers of the Beaufort Group. The first faunal zone of the Early Triassic in South Africa (fig.6) is named for the small dicynodont Lystrosaurus, a widely dispersed genus who is by far the most commonly found terrestial animal during the Early Triassic. Lystrosaurus will be further discussed in later sections on the Triassic period.

              8. Cynognathus Assemblage Zone (Early Triassic, 247-242 mya). This zone, located on the inner edge of the Beaufort Group (fig.6), will likewise be discussed in later sections.

.

Late Permian dicynodonts: variation and distribution

           Dicynodonts, well represented in the three Late Permian faunal zones described above for the Karoo Basin of South Africa, and surviving through into the Early Triassic, were a widely distributed group in the southern land areas of Gondwanaland. First appearing in the Middle Permian zones of the Beaufort Group, they became the most successful and abundant land vertebrates of the Late Permian, radiating into a varety of niches in widespread regions as large, medium-sized, and small herbivores, and short-limbed burrowers. 

        They were first reported in 1845 by the geologist Andrew Geddes Bain from fossils he found during surveys of South Africa. In a letter published in Transactions of the Geological Society of London,  Bain called them "bidentals" for their two prominent canines or tusks (Bain 1845). The same year, Richard Owen named two South African species of dicynodonts as Dicynodon lacerticeps and Dicynodon bainii, later proposing the more general group name of Dicynodontia in 1860.

        As members of the Synapsids, the single temporal opening of dicynodonts is greatly enlarged, supporting very powerful jaw muscles. Their dentition, while somewhat variable, was usually minimal. Many genera, including the type genus Dicynodon, had a complete absence of teeth, apart from their prominent upper canines (fig.8). Some dicynodonts such as Oudenodon had no teeth at all; others, however, such as Pristerodon, had several post canine teeth for chewing, in addition to large canines. One widely distributed genus, Enthiodon,  had two rows of 9-11 teeth on the palatine bones of its upper jaw, and matching teeth in its dentary or lower jaw bone, as described by Owen (1879). More consistently, the dentary processes of dicynodonts were made up of a horny, keratin covering on both the upper and lower jaws. This formed a beak-like appearance, resembling that of modern turtles.  

            Regarding their reproduction, it remains unknown whether dicynodonts laid eggs, as in reptiles, or had developed vivipary, as in mammals. The bony birth canal area between the pubis and ischium was large enough to allow the birth of live young, but could also have been useful in laying large eggs (King 1990).  Some genera display sexual dimorphism, as shown by larger canine teeth in males.

A distribution map (fig.9) shows the areas in Gondowanaland where fossils of four widespread Late Permian dicynodont genera have been found (Ray 1999). These include Cistecephalus (already discussed as the type genus for a faunal zone), Endothiodon, Pristerodon, and Oudenodon. Based on the overlapping distribution pattern of the widespread conifer Glossopteris these animals probably lived in wooded, well-watered, riverine settings. 

Fig.9:  Late Permian Pangea, showing distribution of four Late Permian dicynodonts (after Ray 1999).

 Endothiodon

 The Late Permian dicynodont Endothiodon (fig.10) was first          described by Owen (1879) for the Karoo region, with a skull and mandible, including 11 teeth located in the palatines. A partial skeleton was later recovered by Broom (1915) at Beaufort West in South Africa, from the Hoedemaker member of Middle Teekloof Formation, dating from the Late Permian Tatarian Age.  Endothiodon augusticeps is one of the species found in South Africa.  Since then, Endothiodon has also been found in Tanzania (Haughton 1932), Zambia (Mazin and King 1991), Mozambique (King 1992), India (Kutty 1970), and Brazil (Boos et al. 2013). 

 Fig.10:  Endothiodon skeleton (after Broom 1915; AMNH) .

            The finding of Endothiodon in Brazil is particularly noteworthy, since it is the first dicynodont to be reported for the Permian of South America. A partial skull and associated lower jaw were discovered in the 1970s in a railway cut at Serra do Cadeado, in the Rio do Rasto Formation of Parana State in Brazil (Barbarena and Araujo 1975). The Rio de Rasto Formation is dated at Ufimian to Early Tatarian age, in the Guadelupian Stage of the Middle Permian. The fossils were initially assigned to the genus Endothiodon Owen, which implied a direct correlation of the Brazil specimen with the taxon from the Beaufort Group in South Africa. 

            Boos et al. (2013) recently reexamined the specimen (FURB PV0226), and have confirmed its original assignment by Barbarena and Araujo to the genus Endothiodon. This is based on several key anatomical features, such as the pineal foramen being located on a boss or small protrusion, a bulbous swellings of the dentaries, a boss situated on the ventral margin of the jugal, an extensive number of dentary teeth, and an upturned pointed beak of the lower jaw (Boos et al. 2013). Further analysis is considered necessary to determine the exact species identification. 

       Another interesting facet of the findings of Boos et al. pertain to biostratigraphical correlations they propose for the tetrapod faunas of the Rio do Rasto Formation. These show similarities with stratigraphic ranges of the Middle and Late Permian of both South Africa and Eastern Europe. As Boos et al. (2013) state, the presence of Endothiodon shows that part of the Rio do Rasto Formation in Brazil can now be correlated with deposits in India, Malawi, Mozambique, South Africa, Tanzania, Zambia, and Zimbabwe.

Pristerodon             
            Pristerodon
was first identified in South Africa, and described by T.H. Huxley in 1868. It has since been found in both India and Tanzania
(Ray 1999). Pristerodon, whose skull and likely associated muscles have been studied in detail, had a skull 4-6 cm long, with a short snout (fig.11). Unlike most other dicynodonts, Pristerodon retained six pointed post-canine teeth on either side of its upper and lower jaws (maxilla and dentary).  The rest of the jaws were made of sharp bone, probably covered with keratin or a horny surface (Benton 2005). Pristerodon lived on vegetation which it snipped off with its horny beak, then passed back to its cheek area for grinding with its  molars before swallowing.
            Pristerodon's jaw was built as a flexible unit, able to slide back and forth while eating. The key chewing muscles included a 1) large lateral external adductor that contracted the jaw, running from the back of the skull (squamosal and quadratojugal bones) to a long ridge on the side of the dentary; 2) another jaw contractor, the medial external adductor that ran inside the zygomatic arch from the parietal bone at the top of the skull, to the top of the dentary; 3) a small pterygoidus muscle that pulled the jaw forward; and 4) the depressor mandibulae, which opened the jaw. This ran from the back of the skull (squamosal bone) to the retroarticular process, the part of the jaw behind the jaw joint (Benton 2005).

 Fig.11:  Pristerodon skull (A) and lower jaw (B), showing teeth (after Benton 2005) 

Oudenodon           

            Oudenodon is the type genus of the family Oudenodontidae, which also includes Cteniosaurus, Tropidostoma, and Rhachiocephalus. The Oudenodontidae family have a deep, thin walled cleft in the lower jaw for anchoring the horny platform. They had a very precise cutting and crushing surface between the beaks, where the rim of the maxilla was pointed downward to fit with the lower jaw symphesis. Oudenodon (fig.12) had no teeth at all. The skull was directed forwards, therefore, adapted to feeding at levels approximately 20-100 cm above the ground (King 1990). This shows a new level of diversification, away from substrate-feeding.

Fig.12:  Oudenodon skull, showing beak-like jaws (Berlin Museum of Natural History)

            Oudenodon was common throughout southern Africa during the Late Permian, and has also been found in India (Ray 1999; Bandyopadhyay 1999), Madagascar, and Zambia. Several known species of Oudenodon include Oudenodon bainii, the type species, and O. grandis from South Africa, and O. luangwensis from Zambia.  Another species, O. sakamenensis, is the only therapsid yet known from Madagascar (Mazin and King 1991). 

Gorgonopsians: the main Late Permian predators

            Gorgonopsians (Gorgon and -ops-, "eye" or "face" = "Gorgon faces"), are the most primitive of the three groups of theriodonts found in the Karoo Basin, the other two being therocephalians and cynodonts. During the Late Permian, the gorgonopsians were a highly successful group, including a total of 25 genera and 41 species, from South Africa, Tanzania, Zambia, and Malawi, as well as from Russia (Sigogneau-Russell 1989; Ivokonenko 2005; Gebauer 2007). 

            In the Beaufort Group of South Africa’s Karoo Basin, the earliest significant finds of gorganopsians are in the Middle Permian, Tapinocephalus Zone (266-260 mya), representing the three genera Gorgonops, Eoarctops, and Galesuchus.  Colbert (1948) proposed Galesuchus as the most primitive of the known gorgonopsians from South Africa.

Fig.13:  Skeleton of Sauroctonus, a Late Permian gorgonopsian found in Russia and Africa  (Gebauer 2007)

            During the Late Permian, gorgonopsians became much more common, with the genera Gorgonops, Eoarctops, Lycaenops, Sauractonus (figs.13,16) Arctognathus,  Njalila, Clelandinaand Rubidgea being found in the Tropidostoma and Cistecephalus Zones dated at 260-255 mya (Rubidge 1999; Smith and Keyser 1995). In the Dicynodont Zone (255-252 mya) occur the latest forms of Gorgonopsians, including Cynosaurus, Prorubidgea, and Rubidgea, the latter with extremely enlarged canines. Gorgonopsians then became extinct as part of the mass extinction at the end of the Permian (251 mya), with the other two theriodont groups continuing into the Triassic (Rubidge et al. 1995; Gebauer 2007).

             Gorgonopsians are thought to have evolved in the Middle Permian (ca. 272-265 mya) from a reptile-like therapsid. The early gorgonopsians were small, dog-sized therapsids. The extinction of dinocephalians around the end of the Middle Permian, however, enabled gorgonopsians to become the dominant Late Permian predators.  Their success as hunters was partly due to their large canines, and to changes in their limbs allowing more efficient movement. Their saber-like teeth, which grew larger in later species, could penetrate the tough, scaly skin of some slow-moving herbivores, including pareiasaurs and other  therapsids. The legs of gorgonopsians, importantly, supported their bodies from below rather than sprawling out to the sides in the reptilian mode, and enabled speed as well as strength. This provided a decisive advantage over animals they hunted, such as the armored pareiasaur Bradysaurus (fig.14).

Fig. 14: Skull of Bradysaurus seen from above and from the side, showing the rough, scaly armor used as a defense against predators (after Smith and Keyser 1995).

          Gorgonopsians were among the largest carnivores of the Late Permian. Most of the known forms, such as the type genus Gorgonops, were 1.5 - 2 meters in length, about the size of lions and other large felids. The largest known genus was Inostrancevia, found in Late Permian deposits in the Northern Dvina Valley in Russia, in the Sokolki assemblage (Efremov 1937; Ivakhnenko 1990). Inostrancevia was over 3 meters long, about the size of the largest modern bears, with a skull length of 45 cm, and saber-like canine teeth up to 12 cm long.          

          The mammal-like traits of gorgonopsians include differentiated or heterodont teeth, with incisors, canines, and cheek teeth or molars; a large, fully developed temporal fenestra for strong jaw muscle attachment; and vertically-aligned rear legs for faster movement. Gorgonopsids also had a vaulted palate that may have facilitated breathing while holding prey in their jaws; and early stages of the hearing ossicles that later developed in the mammalian middle ear.

             The type species, Gorgonops torvus, was first identified on the basis of an incomplete and flattened skull found in South Africa at Mildenhalls, Fort Beaufort. Among the first of the therapsids to be described (Owen 1876), G. torvus was also used as the type genus both for the Gorgonopsidae family (Lydekker 1890),and the suborder Gorgonopsia (Seeley 1895).  Seeley's grouping, however, was based on a mistaken notion that gorgonopsians lacked temporal openings for muscle attachments, as opposed to theriodonta (Gebauer 2007). This was corrected by Broom in 1913, who established the subfamily of gorgonopsia as distinct from the theriocephalians (Broom 1913b). From then onward, continued fossil discoveries in South Africa, Tanzania, Zambia, Malawi, and Russia, and South America are described in a long series of publications, beginning with Broom (1912, 1913a, 1925, 1930, 1932, 1940) and Watson (1914, 1921), along with numerous other authors cited below.

          The type genus Gorgonops, like most members of the family gorgonopsidae, has a body length of 1.2-2 meters and a skull length of 22-35 cm. Its formidable array of teeth included the large canines, which had serrated edges, and five incisors. Besides G. torvus, among the earlier species to be described were the larger Gorgonops whaitsi (Broom, 1912), found in the Tropidostoma and Cistecephalus Zones in the Beaufort West area. Sigogneau-Russell (1989) considered this a more primitive form than G. torvus. A third and still larger species, which may be descended from G. whaitsi, is Gorgonops longifrons (Haughton 1915),  known from an incomplete and flattened skull about 35 cm long, with larger eye orbits and a longer snout. 

Fig.15:  Skeleton of Lycaenops, a Late Permian gorgonopsian found in the Beaufort Group (photo: after Gebauer 2007)

           Lycaenops ("Wolf-face"), found in the Late Permian Endothiodon Zone, was another medium-sized gorgonopsian, with more enlarged canines. The first named species of this genus was Lycaenops ornatus (fig.15), whose type specimen was a nearly complete skeleton found by Broom in 1920 on a weathering shale slope near a railway line, two miles south of the railway station at  Biejespoort, South Africa. Stratigraphically, it was at the very top of the Endothiodon zone, within the Teekloof Formation of the Karoo Basin.  Some carpal bones of the hands and feet had weathered away along with most of the tail and cervical region, but otherwise the skeleton was intact, a rarity for Gorgonopsia fossils (Broom 1930, pp.349-50; Colbert 1948).                               
            Broom published descriptions of L. ornatus in 1925, 1930, and 1932. The fossil, obtained by the American Museum of Natural History
(AMNH No. 2240), was cleaned of all stone matrix and fully restored, with a number of new new details found in the skull and skeleton reported by Colbert (1948). This provided insights for more efficient classification of gorgonopsians. The species Symnognathus angusticeps, previously described by Broom (1915) was later reclassified as Lycaenops angusticeps. It has a longer, more narrow snout than L. ornatus, but is otherwise quite similar in its skull anatomy.

            Regarding the undoubted identity of gorgonopsians as Late Permian "saber-toothed" carnivores, Colbert observed that the condition of hypertrophy of the canines, accompanied by a deepening of the anterior part of the lower jaw, is well represented in Lycaenops ornatus, but is seen at its most advanced form in the late gorgonopsians Inostrancecia and Rubidgea (Colbert 1948, pp.373-375).  More recently,  Gebauer (2007) has made a detailed comparison between the gorgonopsian Sauroctonus (figs.13,16), first described from the Northern Dvina region in Russia (Tatarinov 1974), and the saber-toothed cat Smilodon fatalis (fig.17), which had similar body lengths of about 2 meters. 

Fig.16:  Skull and anterior portion of the skeleton of Sauroctonus parringtoni, from the Upper Permian in the Ruhuhu Valley, Tanzania; image reversed (Staatliches Museum für Naturkunde, Karlsruhe, Germany).

            In spite of the fact that they are separated by at least 230 million years, and that Sauroctonus still retained some reptilian features in its lower jaw, skull, and skeleton, while Smilodon was an advanced mammal, there are some intriguingly close functional parallels between these two extinct carnivores, both in teeth and jaw forms, and the musculature of jaw movements used in catching and eating prey.

            The lower  jaws of gorgonopsians, for example, had a thicker front portion than the rear portion, protecting the enlarged canine teeth; a similar function was performed by bone flanges of saber-toothed cats. After a detailed analysis, Gebauer concludes that "it can be well imagined that Sauroctonus in all probability occupied the same ecological niche [as] its mammalian relative 230 million years later"  (Gebauer 2007). 

Fig.17:  Skeleton of the saber-toothed cat Smilodon fatalis (after Gebaur 2007).

Gorgonopsians in Tanzania, Malawi, and Zambia

            Haughton (1927) later described  two gorgonopsian genera found in Malawi, Gorgonops and Aelurognathus (fig.18), the latter genus previously named by Owen (1881).  Boonstra (1934), based on analysis of post-cranial anatomy, described two species of the genus Aleruognathus, and one of the genus Arctognathus.         

            Boonstra (1953) later described gorgonopsians from Tanzania, as did Parrington (1955, 1972), including those found by the German engineer Nowack in the 1930s. Identified taxa from Tanzania now include
Aloposaurus, Leontocephalus, Dinogorgan, Sauroctonius, and possibly Arctognathus (Gebauer 2007).
Including the Tanzanian materials,  Kemp (1982, 2005) has made detailed comparisons of gorgonopsian skull and skeletal anatomy with other synapsids.  More recently, Gebauer (2007) reanalyzed one of the best preserved of the Tanzanian specimens, a nearly complete skeleton originally described as Aelurognathus, now identified as a variant of the Russian genus Sauroctonius (Tatarinov 1974). 

Fig.18:  Skull of Aelurognathus, a Late Permian gorgonopsian found iin Tanzania, Malwai, and Zambia, as well as South Africa.

          From Zambia, specimens have been identified as Aelurognathus by Sigogeneu (1970). In general for the gorgonopsians, many problems in redundant classifications (such as addressed by Colbert in 1948) were sorted out by Sigogneau-Russell (1989), although some difficulties remain, given the number of gorgonopsian taxa and the wide geographical range of their presence.  

         The most primitive known gorgonopsians are pelycosaur-like forms in Late Permian Russia, including Biarmosuchus and Eotitanosuchus from Ezhevo. They are contemporary with more advanced forms from South Africa, and thus can provide no simple answers to the origins of gorgonopsians. Other, more derived Russian forms include Sauroctonius, Viatkogorgon, and Suchogorgon (Tatarinov 1974, 1999, 2000; Ivakhnenko 1990). These and other Russian taxa are discussed in the following section.

 

 .

 Glossary        

.

 References:

Amson, E. and M. Laurin  2011 . "On the affinities of Tetraceratops insignis, an Early Permian synapsid." Acta Palaeontologica Polonica 56: 301–312.

Anderson, J.M., and A.R. Cruikshank 1978. "The Biostratigraphy of the Permian and Triassic. Part 5: A review of the classification and distribution of Permo-Triassic tetrapods." Palaeontologia Africana 21, pp. 15-44.

Bain, A.G.  1845. "On the discovery of fossil remains of bidental and other reptiles in South Africa". Transactions of the Geological Society of London, pp. 53–59.

Bandyopadhyay, S. 1999. "Gondwana Vertebrate Faunas of India." PINSA 65, pp. 285-313.

Barberena, M.C., D.C. Araujo, and E.L. Lavina 1985. "The evidence for close paleofaunistic affinity between South America and Africa as indicated by Late Permian and Early Triassic tetrapods." In Ulbick, H.Y. and A. Rocha-Campos (eds.), Gondwana Proceedings, 7th International Gondwana Symposium, San Pablo, pp. 454-467.

Barry, T.H. 1974. "A new dicynodont ancestor from the Upper Ecca." Annals of the South African Museum 64, pp.117-136.

Boonstra, L.D. 1938. "A report of some Karoo reptiles from the Luangwa Valley, Northern Rhodesia" Quaternary Journal of the Geological Society of London 94, pp. 371-384.

Boonstra, L.D. 1953 "A report on a collection of fossil reptilian bones from Tanganyika Territory." Ann. S. Afr. Mus.,42,  pp.5-18

Boonstra, L.D. 1954."The cranial structure of the Titanosuchian Anteosaurus."  Ann. S. Afr. Mus., 42, pp.108-148.

Boonstra, L.D. 1969."The fauna of the Tapinocephalus Zone (Beaufort Beds of the Karoo)." Ann S. Afr. Mus. 56: 1-73.

Boos, A. D. S.; Schultz, C. L.; Vega, C. S.; Aumond, J. S. J.  2013. "On the presence of the Late Permian dicynodont Endothiodon in Brazil". In Angielczyk, Kenneth.

Broom, R.  1912.  "On some new fossil reptiles from the Permian and Triassic beds of South Africa."  Proc. zool. Soc. London,  pp. 859-876.

Broom, R.  1913a. "On a nearly perfect skull of a new species of the Gorgonopsia."  Ann. S. Afr. Mus., 12, pp. 8-10.

Broom, R  1913b. "On the Gorgonopsia, a sub-order of the mammal-like reptiles."  Proc. zool. Soc. London, 1, p.. 225

Broom, R. 1925.  "On some carnivorous therapsids."  Rec. Albany Mus., 3. pp. 309-326.

Broom, R. 1930. "On the structure of the mammal-like reptiles of the sub-order Gorgonopsia"   Phil. Trans. Roy. Soc.

London, Ser. B, 218, pp. 345-371.

Broom, R. 1940. "On some new genera and species of fossil reptiles from the Karroo beds of Graaff-Reinet." Ann. Transvaal Mus., 20, pp. 157-192.

Chowdhury. T.R. 1970. "Two new dicynodonts from the Triassic Yerrapalli Formation of central India."  Palaeontology 13, pp.133-144.

Colbert, E.H.  1948. "The mammal-like reptile Lycaenops."  Bull. Am. Mus. Nat. Hist., 89, pp. 357-404.

Efremov, I.A. 1937.  "On the stratigraphic division of the continental Permian and Triassic USSR on the basis of the terrestrial vertebrate fauna." Dokl. Akad. Nauk SSSR, Nov. Ser., 16(2), pp.125-132.

Gebauer, E.V.I.  2007 . "Phylogeny and evolution of the Gorgonopsia with a special reference to the skull and skeleton of GPIT/RE/7113 ('Aelurognathus?' parringtoni)"  Ph.D. thesis, Tübingen: Eberhard-Karls Universität Tübingen. pp. 1–316.

Gregory, W.T.  1926. "The skeleton of Moschops capensis Broom, a dinocephalian reptile from the Permian of South Africa." Bull. Amer. Mus. Nat. Hist. 56: 179-251

Haughton, S.H.  1915. "Investigations in South African fossil reptiles and Amphibia, 7.  On some new gorgonopsians."  Ann. S. Afr. Mus.,12: 82-90.

Haughton, S. H. 1917. "Investigations in South African fossil reptiles and Amphibia. Part 10. Descriptive catalogue of the Dicynodontia." Annals of the South African Museum 12, pp.127-174.

Haughton, S.H.  1927.  "On Karroo vertebrates from Nyasaland."  Trans. Geol. Soc. S. Afr., 29, pp. 69-83. 

Haughton, S. H.  1932 . "On the collection of Karoo vertebrate fossils from Tanganyika Territory."  Quarterly Journal of the Geological Society of London 84, pp. 634-668.

Ivakhnenko, M.F.  1990.  "The late Palaeozoic faunal assemblage of tetrapods from deposits of the basin of the  Mezen River."  Paleont. J. 24, pp.104-112.

Ivakhnenko, M.F.  2005. "Comparative survey of Lower Permian tetrapod faunas of Eastern Europe and South Africa." Paleont. J., 39, pp.66-71.

Jacobs, L. L., Winkler, D. A., Newman, K. D., Gomani, E. M. & Deino, A., 2005, "Therapsids from the Permian Chiweta Beds and the age of the Karoo Supergroup in Malawi." Palaeontologia Electronica. Vol. 8, #1, pp. 28A: 21-23.

Kammerer, C.F. and K.D. Angielczyk 2009. "A proposed higher taxonomy of anomodont therapsids". Zootaxa 20, pp. 1–24

Kemp, T.S. 1982.  Mammal-like reptiles and the origin of mammals. Academic Press., London, xiv + 363 pp.

Kemp, T.S. 2005.  The Origin and Evolution of Mammals. Oxford University Press, pp.331.

Keyser, A. W.  1975 . "A re-evaluation of the cranial morphology and systematics of some tuskless Anomodontia." Memoir of the Geological Society of South Africa 67, pp.1-110.

King, G.M. 1988.  "Anomodontia. Encyclopedia of Paleoherpetology, Part 17c, Gutsav Fischer Verlag, Stuttgart & New York.

King, G.M. 1990. The Dicynodonts: A Study in Palaeobiology. Chapman & Hall.

King, G.M. 1992. "The paleobiogeography of Permian anomodonts."  Terra Nova 4, pp.633-640.

Kutty, T.S. 1972. "Permian reptile fauna from India." Nature 237, pp.462-463.

 Laurin, M.  1998 . "New data on the cranial anatomy of Lycaenops (Synapsida, Gorgonopsidae), and reflections on the possible presence of streptostyly in gorgonopsians." Journal of Vertebrate Paleontology 18: 765-776.

Mazin, J. M. and King, G. M. 1991. "The first dicynodont from the Late Permian of Malagasy." Palaeontology 34, pp.837–842.

Ochev, V.G., 2001. Zonal stratigraphic correlations of the Upper Permian from the Cis-Urals and South Africa according to tetrapods. 6th European Workshop on Vertebrate Palaeontology - Florence and Montevarchi.

Olson, E. C.  1937.  "The cranial morphology of a new gorgonopsian." J. Geol., 45, pp.511-527.

Owen, R.  1860. "On some reptilian fossils from South Africa." Quaternary Journal of the Geological Assotiation of South Africa 67, pp.1-110.

Owen, R.  1876a.  "Evidences of theriodonts in Permian deposits elsewhere than in South Africa."  Quat. J. Geol. Soc. London, 27, pp.352-363.

Owen, R.  1876b  Descriptive and Illustrated Catalogue of the Fossil Reptilia of South Africa in the Collection of the British Museum. London, British Museum.

Owen, R. 1881. "On the order Theriodontia with a description of a new genus and species (Aelurognathus fel. ow.)." Quat. Jour. Geol. Soc. London, 37, pp. 261-265.

Parrington, F.R. 1955. "On the cranial anatomy of some gorgonopsids and the synapsid middle ear."  Proc. Zool. Soc. London, 125, pp.1-40.

Parrington, F.R. 1974. "A new genus of gorgonopsid from East Africa." Ann. S. Afr. Mus., 64:, pp. 47-52.

Ray, S. 1999. "Permian Reptilian Fauna from the Kundaram Formation, Pranhita-Godavari Valley, India." Journal of African Earth Sciences, 29 (1), pp.211-218.

Rubidge, B.S. 1990 "A new vertebrate biozone at the base of the Beaufort group, Karoo Sequence, South Africa." – Pal. Africana., 27, pp.17-20.

Rubridge, B.S. (ed.) 1995. "Biostratigraphy of the Beaufort Group."  South African Commission on  Stratigraphy, Biostratigraphic Series 1, pp.1-45.

Rubridge, B.S., G.M. King, and P.J. Hancock 1994. "The postcranial skeleton of the earliest dicynodont synapsid Eodicynadon from the Upper Permian of South Africa." Palaeontology 37 (2), pp.397-408.

Seeley, H.G. 1895. "Researches on the structure, organization and classification of the fossil Reptilia. IX, section 1. On the Therosuchia."  Phil. Trans. Roy. Soc. London, B, 185, pp. 987-1018.

Sigogneau D. 1970 "Révision systématique des gorgonopsiens sud-africains."  Cah. Paléont., Paris, XII + 414 pp

Sigogneau-Russell, D., 1989.  "Theriodontia I - Phthinosuchia, Biarmosuchia, Eotitanosuchia, Gorgonopsia" Part 17 B I, Encyclopedia of Paleoherpetology, Gutsav Fischer Verlag, Stuttgart and New York

Simpson, G.G.  1943 . "Mammals and the Nature of Continents". American Journal of Science 241, pp.1–31. 

Smith, R.H.M. and Keyser, A.W. 1995. Biostratigraphy of the Tropidostoma Assemblage Zone. Geological Survey of South Africa, South African Committee for Stratigraphy, Biostratigraphic Series, 1:18-22.

Suess, Eduard  1885-1909. Das Antlitz der Erde [The Face of the Earth]. F. Tempsky, Vienna

Tatarinov, I. P. 1974.  Theriodonts of the USSR.  Tr. Paleont. Inst. Akad. Nauk SSSR, 143: 1-240.

Tatarinov, I. P. 1999.  "New Theriodonts (Reptilia) from the late Permian Fauna of the town of Kotelnich, Kirov region." Paleont. J., 33, pp. 540-546.

Tatarinov, I. P. 2000. "A new Gorgonopid (Reptilia, Theriodontia) from the Upper Permian of the Volgoda Region." Paleont. J., 34: 64-72.

Watson, D.M.S. 1914.  "Notes on some carnivorous reptiles."  Proc. Zool. Soc. London, 4, pp.1021-1038.

Watson, D.M.S. 1921. "The bases of classification of the Theriodonta"  Proc. Zool. Soc. London, 1, pp. 34-98.

Watson, D.M.S. 1948.  "Dicynodon and its allies."  Proc. Zool. Soc. London, 118, pp.823-877.

Wegener, Alfred 1912. "Die Entstehung der Kontinente" [The formation of continents] Geologische Rundschau, Zeitschrift für allgemeine Geologie , dritter Band, pp. 276-292 

Wegener, Alfred 1929 . Die Entstehung der Kontinente und Ozeane [The Origin of Continents and Oceans]  (4th ed.). Braunschweig: Friedrich Vieweg & Sohn Akt.

Wegener, Alfred 1966. The Origin of Continents and Oceans. New York: Dover. (Translated from the 4th revised German edition by John Biram.)

Willem A. J. M. et al.  1928. Theory of Continental Drift: a Symposium of the Origin and Movements of Land-masses of both Inter-Continental and Intra-Continental, as proposed by Alfred WegenerTulsa, OklahomaThe American Association of Petroleum Geologists & London, Thomas Murby & Co.





Athena Review Image Archive™   | Guide to Archaeology on the Internet   |   free trial issue |  subscribe  |  back issues

Main index of Athena Review   |   Subject Index   |   Research Pages   |  Galleries aand Museums  |  Ad rates |  Current issue index


Copyright  ©  1996-2014    Athena Publications, Inc.  (All Rights Reserved).