ISSN 0031-0301, Paleontological Journal, 2008, Vol. 42, No. 9, pp. 859–995. © Pleiades Publishing, Ltd., 2008.

Cranial Morphology and Evolution of Permian Dinomorpha (Eotherapsida) of Eastern Europe M. F. Ivakhnenko Paleontological Institute, Russian Academy of Sciences, Profsoyuznaya ul. 123, Moscow, 117997 Russia e-mail: [email protected] Received July 10, 2007

Abstract—A Revision of the cranial morphology of Middle and Late Permian East European Dinomorpha (Theromorpha, Eotherapsida) resulted in a clarification of the taxonomic relationships within this group. As the system was constructed, four parameters, i.e., morphological, biomorph, and geographical and stratigraphic ranges, were taken into account in each subordinate taxon. The taxon Theromorpha is ranked as class, since it was primarily formed as a pilidosic group (covered with hair), in contrast to the class Reptilia, a pholidosic group (covered with scales). DOI: 10.1134/S0031030108090013 Key words: Theromorpha, Eotherapsida, Dinomorpha, morphology, biomorphs, Permian, eastern Europe.

CONTENTS INTRODUCTION CHAPTER 1. POSITION OF DINOMORPHA IN THE SYSTEM AND DEFINITION CHAPTER 2. MATERIAL AND LOCALITIES CHAPTER 3. COMPARATIVE CRANIAL MORPHOLOGY 1. Skull Roof (ossa calvaria) 2. Cheek Segment (ossa buccalia) 3. Temporal Segment (ossa temporalia) 4. Palatal Segment (ossa palatalia) 5. Lower Jaw (ossa mandibularia) 6. Ethmoidal and Braincase Regions (endocranium) 7. Auditory Structures 8. Dentition 9. Ontogenetic Changes in Skull CHAPTER 4. CRANIAL MORPHOGENESIS CHAPTER 5. ECOBIOMORPHS CHAPTER 6. DISTRIBUTION AND GEOGRAPHICAL RANGES CHAPTER 7. PRINCIPLES OF SYSTEMATICS Superorder Nikkasauria Superorder Gorgodontia Order Dinocephalia Suborder Niaftasuchida Suborder Dinocephalida Order Gorgonopia Suborder Ictidorhinida Suborder Gorgonopida Suborder Estemmenosuchida Superorder Anomodontia Order Ulemicia Order Dicynodontia Suborder Dromasaurida Suborder Dicynodontida CONCLUSIONS ACKNOWLEDGMENTS REFERENCES 859

860 863 865 879 879 887 891 896 904 913 925 929 942 947 965 972 979 980 980 980 981 981 984 984 984 986 987 987 988 988 988 990 991 991

860

IVAKHNENKO

INTRODUCTION In the present study, the name Dinomorpha designates the higher group of Eotherapsida. Eotherapsida along with the sister group Eutherapsida compose the taxon Theromorpha, which is usually regarded as a subclass. In classical schemes, this subclass is included in the class Reptilia, along with the classes Amphibia, Aves, and Mammalia of the superclass Tetrapoda. This scheme was developed more than hundred years ago, taking into account only the morphology of extant groups and using criteria (mostly connected with physiology) difficult to apply to extinct groups. However, the development of a new general classification scheme of tetrapods as a whole is beyond the scope of the present study. In modern literature, Theromorpha is frequently named Synapsida, the term introduced by Osborn (1903) to designate taxa with one (lower) temporal fenestra. However, I think this is inexpedient, since, in my opinion, Theromorpha comprises only a part of the synapsid groups. The present study continues previous research (Ivakhnenko, 2001, 2003c, 2005b). The analysis of the taxonomic composition of Permian specimens from East European localities containing tetrapods has shown that nycteroleteromorph parareptiles and theromorphs are the second most abundant groups after aquatic labyrinthodonts (Ivakhnenko et al., 1997). The natural geosystem formed in eastern Europe in the Middle Permian and continuing for more than 10 m.y. to the onset of the Mesozoic is named the East Europe Placket (Ivakhnenko, 2001). The placket is a natural territorial complex with very special physiographical conditions, which has no analogue in Recent ecosystems. The area of the East European Placket was inhabited by a number of oryctocenotic tetrapod assemblages replacing each other in time. Certain faunal assemblages reconstructed replaced each other in connection with changes in physiographic conditions. In fact, throughout the history of the East European Placket, hygrophilic and mesophilic communities of coastal ecotopes were dominated by theromorphs with reference to both taxonomic and biomorph diversity; they occupied various niches, sometimes participating in hydrophilic communities. Theromorpha of that time include predators and phytophages, ranging in size from approximately rat-sized, to rhinoceros or tiger-sized animals. In the Paleontological Institute of the Russian Academy of Sciences (PIN), a large proportion of the Theromorpha groups of the world are housed; some specimens are perfectly preserved. This material from eastern Europe has long attracted the attention of researchers; the first representatives of Theromorpha were described by Stephan Kutorga in 1838. During the subsequent century, several dozen works were published, describing new taxa of this group from eastern Europe. However, only a few taxa were examined in detail (Efremov, 1954; Orlov, 1958; Tatarinov, 1974; Tchudinov, 1983). These works have contributed

considerably to the study of the group; however, they have a fundamental imperfection, as almost all works dealing with representatives of Dinomorpha from other territories. The point is that they do not include alternative comparative diagnoses of higher taxa (above family level). The diagnoses are replaced by general characteristics or brief descriptions, which are usually composed of the most distinct features of the subordinated taxa. At the same time, comparisons with taxa of the same rank were performed based on isolated and, moreover, different structures. As a result, the diagnosis was determined by the taxa included by an author into a taxon in question rather than the composition of a taxon being determined by its diagnosis. The characteristics obtained were vague and, hence, it was impossible to recognize the principles of assignment of a particular taxon in a certain higher taxon. In particular, this remark concerns Theromorpha as a whole, the only character of which included in consideration was the presence of the temporal fenestra of the synapsid type. However, available data strongly suggest that the position and general structure of the temporal fenestra of tetrapods was preformed in the skull design of some crossopterygians; thus, it could have been formed independently in different groups and should not be taken as the determining character of a lineage. The absence of strict criteria unambiguously determining the position of taxa was a basis for a revision of the principles for the establishment of Theromorpha (Ivakhnenko, 2003c). The definition and systematics of this group are revised based on the reconstructed trend in the formation of the general cranial syndrome. The process passed certain stages of the formation of morphophysiological grade of the group. At the primary tetrapod (prototetrapod) level of a facultatively aquatic adult stage, the elements of the general tetrapod syndrome connected with locomotion, breathing, angustitabular or latitabular skull patterns, and apopareial or synpareial temporal region were developed. At the next (labyrinthodont), obligatory aquatic level, the ancestors of theromorphs developed the structures connected with the sensory system (progressive development of the infradental diverticulum of the seismosensory canal), along with other less important structures. At the amphibian, facultatively terrestrial level, the subapsid incisure developed, as occurred in parallel in a number of other primitive groups; however, in contrast to other groups, the periangular cavity of the lower jaw developed from the infradental diverticulum. At the reptilian, obligatory terrestrial level, the synapsid fenestra opened and the woolly–sweat cover was formed to protect the body against drying under terrestrial conditions. These features determined the syndrome of Theromorpha as apopareial angustitabulars with index structures including the synapsid fenestra, subapsid incisure, and periangular cavity. Therefore, ophiacodontomorph groups of pelycosaurs (Ophiacodontia, Caseida, and Varanopeida) are excluded from Theromorpha. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

Theromorpha are divided into two major groups, Eotherapsida (with a reduced epipteric cavity, braincase of the prootic type, and paraquadratobasal design of the temporal skull region) and Eutherapsida (the epipteric cavity is retained, the braincase is of the alisphenoid type, and the temporal skull region is of the squamosobasal design). This division into two major taxa of Theromorpha probably occurred early in evolution, at the initial period of the formation of the temporal fenestra. This could have been connected with certain differences in the primary development of the temporal fenestra, i.e., the primary reinforcement of the posterior region, with the restriction from behind by the true squamosal (fenestra posterosynapsida), or the primary reinforcement of the anterior region, with the opening at the paraquadrate–squamosal (fenestra anterosynapsida). In this case, it is probable that the two groups of Theromorpha evolved in parallel from a stegal design of the temporal region, that is, from a taxon that should be formally excluded from Theromorpha. It was proposed that Theromorpha as a whole were probably formed at the transition to the reptilian morphophysiological grade (Ivakhnenko, 2003c, pp. 352, 353). This means that they acquired features determining this level independently of other lineages that reached the reptilian grade. This concerns, among other things, such an important structure as the isolating skin layer. The analysis of differences in skin structure and physiology of reptiles and living mammals (Tatarinov, 1976) strongly contradicts the hypothesis of the presence of reptilian isolating scales in mammal ancestors. I believe this conclusion concerns all Theromorpha. Naturally, this was probably an insulating layer composed of primary hair resembling mammalian guard hair, which played the role of a porous cover, regulating evaporation. In my opinion, the presence of the unique cover of hairlike structures explains many features of this group (Ivakhnenko, 2005b, pp. 439, 440). Moreover, the presence of this cover, which apparently retained body heat, could have played an important role in the wide distribution of Theromorpha at the beginning of their evolution. This time probably fell at the end of the Carboniferous and the beginning of the Permian, that is, transition from a glacial epoch to a warm epoch (Chumakov, 2004). Thus, certain advantages of Theromorpha provided the formation of many biomorphs and various taxa. Eotherapsida is divided into two taxa, Sphenacomorpha (without streptostyly of the quadrate–quadratojugal complex) and Dinomorpha (with a streptostylic quadrate–quadratojugal complex). Dinomorpha followed a complex evolutionary route, with the formation of various morphological designs. The material from the Late Permian of eastern Europe is of special interest for a better understanding of these processes; within this large, partially isolated area, physiographical conditions changed repeatedly and essentially during the Middle and Upper Permian. Dinomorpha was represented there by primitive relicts, which came out of refuges during the degression of former communities, PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

861

endemic taxa produced by local evolution, and obviously adventive groups that migrated there as a result of invasions from other areas. Some groups, for example, the large predatory Rubidgea, probably developed in eastern Europe and later migrated into Gondwanaland (Ivakhnenko, 2005a). In fact, the Late Permian was the time of the flourishing and domination of eotherapsids, among which Dinomorpha displayed the greatest diversity of biomorphs, including small insectivores, medium-sized omnivores, and giant phytophages and predators. This complex and unusual Late Permian world, the world of eotherapsids, certainly deserves close study. This was probably the time when the pattern of land tetrapod communities was basically formed; subsequently, it was mostly retained and developed throughout the Mesozoic and, to a certain extent, the Cenozoic. The Anomodontia, one of two Dinomorpha taxa, developed the optimum cranial design for a widely adapted, mostly phytophagous biomorph as early as the beginning of the Late Permian. These animals became the main consumers of the first order in tetrapod communities of the Late Permian and Triassic (and probably survived up to the Early Cretaceous: Thulborn and Turner, 2003). Many studies concerning both the cranial morphology and ecology were devoted to this group, which is relatively homogeneous in morphology. The second Dinomorpha taxon Gorgodontia is more complex in morphology, but was examined to a much lesser extent. However, the role of this group in Late Permian communities was also rather significant. The Gorgodontia are represented by both phytophagous and predatory biomorphs in the dominant and subdominant blocks of communities. In some lineages, primitive taxa persisted for a long time, while others rapidly evolved and specialized. Dinomorpha is certainly related in origin to primitive Sphenacomorpha from the Upper Pennsylvanian– Lower Permian. Sphenacomorpha was relatively thoroughly examined in morphology; however, this concerns only high-specialized taxa from the Lower Permian of North America. Primitive taxa that could have been the initial group for Dinomorpha (probably related to the so-called Haptodontidae from the Lower Permian of western Europe) were described, but poorly examined. The earliest known Dinomorpha appeared in the Middle Permian of eastern Europe and were represented there by all major lineages, which evolved later. Consequently, the formation and initial differentiation of this group occurred much earlier and in a different territory because, in the Early Permian, eastern Europe was a sea basin. In the Middle–Late Permian and Triassic, Dinomorpha was widespread in all continents, including Antarctica. Further study of Dinomorpha was impeded by relative scarcity of morphological data on the majority of representatives. Therefore, the relatively well-preserved material of one of groups, the Gorgonopidae,

862

IVAKHNENKO

was examined in detail in the previous study (Ivakhnenko, 2005b). Note that interpretation of many structures was extremely difficult and the results obtained were not unequivocal. This is connected with the fact that, in morphology, representatives of Dinomorpha in some sense are only similar to Mammalia. However, for example, in the structure of the braincase, they differ considerably from the thoroughly examined therian mammals. At the same time, monotremes, which are much more similar to Dinomorpha in the structure of this region, are unsatisfactorily described in the literature. In addition, Dinomorpha shows many distinctive features, which have no analogue in other tetrapod groups. This concerns, for example, the complex surface sculpture of a number of membrane bones and some structural features of the lower jaw. An insight into many complex structures is facilitated by the study of structural variations in various taxa. However, some skull regions of Dinomorpha have not been described. Therefore, it makes sense to describe certain structures (such as the ethmoid region and nasal capsule) even if they are difficult to interpret. It is possible to consider the schemes of the major evolutionary trends in cranial morphogenesis of Dinomorpha as a major result of such studies. However, the cranial structure of a particular animal is closely connected with its biomorph. The head is the region where the main organs of sense and the jaw apparatus are located, that is, these regions contribute considerably to the reconstruction of the biomorph, the life form of animal. The jaw apparatus and dentition are of special interest in this respect. In Dinomorpha, these characteristics are still incompletely understood. At the same time, evolution of any primitive tetrapod group is a regular process of ecobiomorph evolution. In the case of Late Paleozoic tetrapods, including Dinomorpha, this process was closely connected with changes in physiographic conditions of low, placket ecotopes (Ivakhnenko, 2006). In the initial pioneer communities inhabiting the coasts of epeiric basins, primitive representatives of almost all large tetrapod groups were formed. Most xerophilous (diapsid reptiles) and rheophilic (anurans) derivatives left for the presently poorly understood ecotopes of plakors and almost disappeared for some time from the paleontological record. Hydrophilous groups (aquatic batrachomorphs, anthracosauromorphs, and discosauriscomorphs), hygrophiles, and mesophiles (eotherapsids and amphibiotic nycteroleteromorphs) formed primitive oligobiomorph megacommunities of coastal lowlands. These megacommunities consist of dominant and subdominant blocks, which only facultatively interacted with each other. At the end of the Permian, in extensive placket plains, higher Dinomorpha (trophically connected phytophages and predators) formed a polybiomorph community, with smoothed distinctions between dominant and subdominant blocks, which was similar in structure to modern mammalian communities. The sharp global climatic changes at the Permian–Triassic

boundary resulted in a degression of the polybiomorphic community, smoothing distinctions between plakors and plackets, and penetration of plakor xerophiles (primitive archosaurs) into near-water communities. This event meant the end of the placket history of Permian communities and the onset of the formation of Triassic plakor plains in the same areas. Plakor plains form the basis of modern ecotopes of terrestrial tetrapods, and transition to them marked the establishment of Meso-Cenozoic communities, which resemble in structure modern communities. These conclusions are tightly related to the concept of morphophysiological grade and ecobiomorphs of tetrapod groups included in communities. It is evident that biomorphs and, the more so, physiological features of extinct groups are reconstructed only tentatively. Such conclusions are not incontrovertible, but give a more or less probable picture depending on the understanding of available material; thus, they should be tested and corrected in future studies. The evolution of any extensive group, including Dinomorpha, is determined by closely related processes of its morphological and biomorphic changes in time in a particular territory. Therefore, the basic parameters determining the position of each taxon in the group are the morphological syndrome, biomorph, geographical range, and the time of existence. Consequently, research comprises the following tasks: (1) Comparative morphological examination of the group and construction of definitions based on syndromes of taxa of all levels. As a result, it is possible to construct the scheme of morphogenesis of the group. (2) The determination of the general morphophysiological grade of the group and patterns of its formation, the reconstruction of megabiomorphs of large taxa and ecobiomorphs of particular taxa. The reconstruction of processes of biomorph evolution in the group examined. (3) The reconstruction of geographical ranges of taxa and revelation of migration routes of adventive components of communities. (4) The determination of the time of existence of taxa of all levels. Certainly, it is only possible to obtain approximate stratigraphic ranges determined by relative ages of available finds. Regarding extinct groups, the temporal range shows that a particular taxon appeared not later and disappeared not earlier than these limits. Hence, true moments of appearance or disappearance of taxa are reconstructed more or less precisely. Thus, the position of each taxon in phylogenetic tree of the groups is determined by four coordinates. Therefore, graphic construction of such a tree is almost impossible, particularly, taking into account constant lack of sufficient morphological data, which is caused by incompleteness of the paleontological record and the tentative nature of all data on biomorphs, since they are based on reconstructions, and on the time of existence of particular taxa. It is only possible to construct PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

approximate schemes, which are constantly corrected by new data. Certainly, this study does not provide an exhaustive account, particularly in the case of a long persisting, extensive, and diverse group such as the Late Paleozoic Dinomorpha. It is only possible to consider some aspects, which are interpretable at the present state of knowledge. It is also important to discuss available data even when they are difficult to interpret because of the absence of comparative data. Otherwise, they will remain uncertain and dead. Further studies and new finds will expand, improve, and develop present-day conclusions. Sooner or later, this will result in the development of a general picture of the evolution of Dinomorpha, a group with diverse morphology and biomorphs, which persisted for a long time and played a dominant role in tetrapod communities at the end of the Paleozoic and the beginning of the Mesozoic. CHAPTER 1. POSITION OF DINOMORPHA IN THE SYSTEM AND DEFINITION Dinomorpha is commonly considered as a group of Theromorpha. The taxon Theromorpha was established by Cope (1878) and included the orders Pelycosauria and Anomodontia. The higher Theromorpha, Anomodontia Owen, 1859, were divided by Seeley (1894) into a number of orders: Theriodontia, Gorgonopsia, Dinocephalia, Deuterosauria, Endothiodontia, and Dicynodontia. Subsequently, the number of taxa was reduced to three (Broom, 1905), i.e., Therocephalia, Theriodontia (Cynodontia: Broom, 1906), and Anomodontia. Watson (1914) included Gorgonopia, Therocephalia, and Cynodontia in Theriodontia and opposed them to Dinocephalia. The widely known scheme proposed by Huene (1948, 1956) only confused the situation, since a number of obviously unrelated groups, such as Mesosauria, Pelycosauria, Therapsida, Placodontia, Sauropterygia, and Protorosauria, were combined under the name Theromorphoidea. The order Therapsida was divided without comment into the suborders Dromasauria, Dinocephalia, Anomodontia, Gorgonopsia, Rubidginia, Therocephalia, Bauriamorpha, Cynodontia, and Ictidosauria. Romer (1956; Watson and Romer, 1956) provided a more consistent scheme; the order Therapsida contained two suborders, Theriodontia and Anomodontia. Gorgonopia was regarded as a primitive group of Theriodontia, while Anomodontia included both Dicynodontia and Dinocephalia. Boonstra (1963a, 1963b) adhered to almost the same scheme, proposing the close affinity of primitive Dinocephalia and Gorgonopia, which were united in the suborder Eotitanosuchia, including both Biarmosuchus (Eotitanosuchidae) and Phthinosuchus (Phthinosuchidae). Subsequently, it was correctly proposed to place these groups close to Dinocephalia and Gorgonopia, respectively (see Fundamentals of Paleontology…, 1964; Vjushkov, 1964; Tatarinov, 1964; Tchudinov, 1964a). However, the separation of Anomodontia from PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

863

these groups and inclusion of Gorgonopia in Theriodontia seems incorrect. Perhaps, Tatarinov (1974) was the first to show that Gorgonopia should be opposed to other Theriodontia, as he considered the orders Phthinosuchia, Gorgonopia, and Eutheriodontia and subsequently united Phthinosuchia and Gorgonopia (Tatarinov, 1976). Thus, the morphological affinity of Dinocephalia and Anomodontia was accepted by many researchers (for example, Kemp, 1988; King, 1988; etc.). The last question to be resolved was whether Gorgonopia should be placed close to Eutheriodontia sensu Tatarinov or to Dinocephalia and Anomodontia. The last variant was implied by the well-supported assignment of primitive representatives of Gorgonopia and Dinocephalia to Eotheriodontia (Olson, 1962) or Eotitanosuchia (Boonstra, 1963a, 1963b). The study of the skull structure (Ivakhnenko, 1999, 2000a) of Gorgonopidae, Phthinosuchidae, and Estemmenosuchidae has shown morphological affinity of these groups and opposition to Dinocephalia. This conclusion was based on the structure of the temporal region. Subsequent studies (Ivakhnenko, 2001, 2002a) have shown that the so-called primitive therapsids (Gorgonopia, Dinocephalia, and Anomodontia) combined with Sphenacodontia compose a single group (Eotheriodontia), which is opposed to higher therapsids (Eutheriodontia, including Scaloposauria, Therocephalia, and Cynodontia). However, Eotheriodontia is a sister group, with a very early independent evolutionary trend, rather than a primitive, morphologically initial group for the Eutheriodontia. This important conclusion was based on two parameters, which probably correlate. The adductor fossa of “Eotheriodontia” extends medially to the prootic wall of the braincase. In “Eutheriodontia,” the cavity is retained between the adductor fossa and braincase (cavum epiptericum, a rudiment of the spiracular cavity); medially, the adductor fossa is bordered by the epipterygoid region of the ascending plate of the palatoquadrate. Thus, this important parameter of “Eotheriodontia” is apomorphic relative to the cranial design of “Eutheriodontia”; hence, the first group is more specialized. In addition, it is likely that the squamosal is not homologous in these two groups. In Sphenacodontia, Gorgonopia, Dinocephalia, and Anomodontia, this bone is homologous to the squamosum of, for example, Lepidosauria, and corresponds to the paraquadratum (Gaupp, 1906). This bone was probably formed below the level of the crista parotica and, hence, as the supratemporal was reduced, it spread under the parietal. In addition, in “Eotheriodontia,” it forms the upper border of the temporal fenestra. This structure of the temporal region was designated the paraquadratobasal design. In “Eutheriodontia,” the squamosum is obviously homologous with the same bone of mammals and, probably, with the supratemporal of Lepidosauria (Gaupp, 1906). In this case, the bone is initiated above the level of the crista parotica and, hence, overlies the parietal. In “Eutheriodontia,” it does not adjoin the postorbital; hence, the upper border

864

IVAKHNENKO

of the temporal fenestra is formed by the parietal. This structure was designated the squamosobasal design. In the subsequent study, Ivakhnenko (2003c) composed the definitions of higher taxa of Theromorpha, which were determined by respective syndromes. In this work, Theromorpha were ranked as a subclass, which possibly corresponded to a lineage, with the following definition: (1) Angustitabularia. The defining character is the presence of contact between the tabular and parietal bones. The syndrome is based on the presence of separate exoskeletons of the intercapsule roof and the roof of the occipital ring, since the tectum synoticum is positioned under the posterior part of the parietal, while the tectum posterius is completely overlain by the postparietal (Ivakhnenko, 1984). (2) Apopareialia. The defining character is the absence of a wide flange of the squamosal, which underlies the supratemporal; a primitive design is retained, with sutureless contact between the bones of the skull roof and cheek region in the position of the spiracular fissure (Ivakhnenko, 1989). (3) Subapsida. The defining character is the absence of contact between the jugal and quadratojugal, which is connected with a primitive fenestration of the skull, the formation of the subapsid fenestra (or incisure) in the region of contact between the jugal, squamosal, and quadratojugal. In Theromorpha, the subapsid skull pattern is determined by the presence of the subapsid incisure, which is manifested in the absence of contact between the jugal and quadratojugal. (4) Synapsida. The defining character is the presence of the synapsid (temporal) fenestra in the region of rudimentary spiracular fissure between the roof and cheek (Kemp, 1969). The fenestra is located below the caudal process of the postorbital, the zygomatic arch is formed by the jugal. (5) Periangularia. The defining character is the presence of a cavity on the external surface of the angular bone of the lower jaw. The cavity is covered externally by a thin bony outgrowth (ala angularis), which continues the external crest (lamina reflexa). The cavity was probably preformed by the inherited bony groove of the infradental seismosensory canal and was connected with the system of sound conduction. The subclass Theromorpha was divided into two infraclasses, the Eutherapsida (= Eutheriodontia sensu Tatarinov, 1974; = Therosauria sensu Kemp, 1982) and Eotherapsida. The infraclass Eutherapsida is characterized by the preservation of the epipteric cavity and by the squamosobasal design of the temporal region. The epipteric cavity separates the lateral wall of the braincase from the medial wall of the adductor fossa, which consists of an expanded epipterygoid (= alisphenoid) and a special flange of the squamosal. The squamosobasal design is determined by the presence of the squamosal, which lies

above the crista parotica and overlies dorsally the parietal. The postorbital lacks contact with the squamosal. The infraclass Eotherapsida is characterized by a reduced epipteric cavity and by the paraquadratobasal design of the temporal region. The epipterygoid is thin, the medial wall of the adductor fossa is the lateral wall of the braincase, i.e., the anterior plate of the prootic and united vertical block (periotica) of fused tectum posterius, tectum synoticum, and opisthoticum. The paraquadratobasal design is determined by the presence of the squamosal, which lies below the crista parotica and underlies dorsally the parietal. The caudal process of the postorbital comes into contact with the squamosal. The infraclass Eotherapsida included two superorders, the Sphenacomorpha and Dinomorpha; with reference to morphology, it is possible to regard them as ancestors and descendants. Sphenacomorpha are presently known from the Late Carboniferous and Early Permian; some representatives have been examined relatively thoroughly (Romer and Price, 1940). Unfortunately, only large-sized taxa from the Early Permian of North America have been studied; they are probably endemic to this region and only remotely related to Dinomorpha. Probably, the most remarkable feature of the cranial structure of these animals, which sharply distinguishes them from Dinomorpha is a very large volume of the part of the adductor fossa that is morphologically located below the level of the crista parotica. In Dinomorpha, the part of the adductor fossa above the level of the crista parotica is much more voluminous than the part below this level. However, this feature may be connected with a distinct trend in the improvement of the jaw musculature (which is caused by the large size of these animals) rather than with the primitive state of Sphenacomorpha. The analysis of skull morphology in North American taxa has shown that the major distinction between Sphenacomorpha and Dinomorpha concerns the improvement of the jaw muscles. In primitive groups of the amphibian morphophysiological grade, the main weight of adductors is usually located in the lower part of the temporal cavity. This part (subotic region) is below the level of the crista parotica (or processus paroccipitalis, if it is present). Therefore, as the musculature is strengthened, the subapsid fenestra or incisure could have been formed in the integument of this region. In the evolution of groups of reptilian morphophysiological grade, the muscles were strengthened mostly in the region above the level of the crista parotica (supraotic region); this resulted in the development of the upper temporal fenestra. In Sphenacomorpha, the subotic and supraotic regions are still almost equal in volume, whereas, in Dinomorpha, the supraotic region is much larger. A reliable criterion distinguishing these groups is the presence or absence of a suture between the squamosal and quadratojugal (Ivakhnenko, 2003c, pp. 357, 358). In each group, the quadratojugal is combined with the quadrate in a single complex (QQJ-comPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

plex); however, in Sphenacomorpha, it is connected throughout its vertical extent by a serrated suture with the medial margin of the squamosal (astreptostylic construction). In Dinomorpha, the squamosal lacks a lateral sutural connection with the quadratojugal. The QQJ-complex is located in a special depression in the occipital plate of the squamosal (streptostylic construction). It was shown that almost all large-sized Dinomorpha have got rid of streptostyly by different means (Ivakhnenko, 2003c); therefore, it is possible to propose that this design originated in small primitive, most likely, insectivorous taxa to facilitate manipulation of mobile prey. Ancestors of Dinomorpha were probably small, primitive, insectivorous animals. Dinomorpha were considered to comprise two orders (Gorgodontia and Anomodontia) distinguished by the development of the subapsid incisure (Ivakhnenko, 2003c). Gorgodontia was divided into the suborders Dinocephalia (with the infraorders Eotitanosuchida, Tapinocephalida, and Titanosuchida) and Gorgonopia (infraorders Gorgonopida and Estemmenosuchida). Anomodontia was divided into the suborders Ulemicia and Dicynodontia. The major distinctive parameters concern the structure and function of the jaw apparatus and dentition. However, the study of skull morphology in a number of representatives of Gorgonopidae (Ivakhnenko, 2005b) has shown that many important cranial features of various Dinomorpha were not taken into account as the scheme was developed, because they had not been described. This was the basis for the present study; as a result, it has become evident that the taxonomic system requires a significant revision. Certainly, it is impossible to perform in one work a thorough examination of the skull morphology of all representatives of the superorder (even those from eastern Europe) that would have been as thorough as the previous study of Gorgonopidae. However, even a less profound study allows essential correction of the taxonomic scheme (see below). For this purpose, representatives of the following families were examined: Nikkasauridae Ivachnenko, 2000 (Nikkasaurus Ivachnenko, 2000; Reiszia Ivachnenko, 2000); Microuraniidae Ivachnenko, 1995 (Microurania Ivachnenko, 1995); Niaftasuchidae Ivachnenko, 1990 (Niaftasuchus Ivachnenko, 1990); Ulemosauridae Ivachnenko, 1994 (Ulemosaurus Riabinin, 1938); Deuterosauridae Seeley, 1894 (Deuterosaurus Eichwald, 1846); Alrausuchidae Ivachnenko, fam. nov. (Alrausuchus Ivachnenko gen. nov.); Eotitanosuchidae Tchudinov, 1960 (Biarmosuchus Tchudinov, 1960); Archaeosyodontidae Ivachnenko, fam. nov. (Microsyodon Ivachnenko, 1995; Archaeosyodon Tchudinov, 1960); Syodontidae Ivachnenko, 1994 (Syodon Kutorga, 1838); Anteosauridae Boonstra, 1954 (Titanophoneus Efremov, 1938); Phthinosuchidae Efremov, 1954 (Dinosaurus Fischer, 1847; Admetophoneus Efremov, 1954; Kamagorgon Tatarinov, 1999; Viatkogorgon Tatarinov, 1999); Rubidgeidae Broom, 1938 (Leogorgon Ivachnenko, 2003); Inostranceviidae Huene, 1948 (InosPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

865

trancevia Amalitzky, 1922); Gorgonopidae Lydekker, 1890 (Pravoslavlevia Vjuschkov, 1953; Sauroctonus Bystrov, 1955; Suchogorgon Tatarinov, 2000); Ictidorhinidae Broom, 1932 (Biarmosuchoides Tverdochlebova et Ivachnenko, 1994; Ustia Ivachnenko, 2003); Burnetiidae Broom, 1923 (Proburnetia Tatarinov, 1968; Niuksenitia Tatarinov, 1977); Rhopalodontidae Seeley, 1894 (Rhopalodon Fischer, 1841; Phthinosaurus Efremov, 1940; Parabradysaurus Efremov, 1954); Estemmenosuchidae Tchudinov, 1960 (Estemmenosuchus Tchudinov, 1960); Venyukoviidae Efremov, 1940 (Venyukovia Amalitzky, 1922; Otsheria Tchudinov, 1960); Ulemicidae Ivachnenko, 1996 (Ulemica Ivachnenko, 1996); Galeopidae Broom, 1912 (Suminia Ivachnenko, 1994); Pristerodontidae Toerien, 1953 (Australobarbarus Kurkin, 2000); and Dicynodontidae Owen, 1859 (Idelesaurus Kurkin, 2006; Dicynodon Owen, 1845; Vivaxosaurus Kalandadze et Kurkin, 2000; Delectosaurus Kurkin, 2001; Elph Kurkin, 1999). CHAPTER 2. MATERIAL AND LOCALITIES The material examined in the present study is stored in the following institutions: (PIN) Paleontological Institute of the Russian Academy of Sciences, Moscow; (SGU) Saratov State University, Saratov (presently, the specimens are stored in PIN); and (TsNIGR Museum) Chernyshev Central Research Geological Museum, St. Petersburg. Only the collection numbers of specimens examined morphologically by the author of the present study are provided. Collection PIN, no. 48 (unknown locality) Russian Federation, Orenburg Region, probably Kargala River mines; Biarmian Series (Middle Permian), Urzhumian Stage; Ocher Faunal Assemblage, Ocher Faunal Subassemblage. Taphonomic association: Venyukoviidae Efremov, 1940: Venyukovia prima Amalitzky, 1922. Material examined: Venyukovia prima Amalitzky, 1922. Specimen PIN, no. 48/1, left dentary (Figs. 53a, 53b, 67a); Amalitzky, 1922, p. 10, text-fig. 14 (without no.); Efremov, 1940a, pp. 57, 58, pl. VIII, fig. 4 (Venjukovia prima, err.); 1954, p. 220, pl. XXVIII, fig. 3 (Venjukovia prima, err.); Tchudinov, 1983, text-fig. 62 (lectotype); Ivakhnenko, 1996, pp. 76–79, text-figs. 1c, 1g; 2001, text-fig. 36b; 2003c, text-fig. 39; Ivakhnenko et al., 1997, pl. 89, fig. 1. Collection PIN, no. 156 (Semin Ovrag locality) Russian Federation, Tatarstan, Tetyushskii District; Tatarian Series (Upper Permian), Upper Severodvinian

866

IVAKHNENKO

Substage, Severodvinian Horizon; Sokolki Faunal Assemblage, Ilinskoe Faunal Subassemblage. Taphonomic association: Dvinosauridae Amalitzky, 1921: Dvinosaurus primus Amalitzky, 1921; Kotlassiidae Romer, 1934: Microphon exiguus Ivachnenko, 1983; Pareiasauridae Seeley, 1888: Proelginia permiana Hartmann-Weinberg, 1937; Chroniosuchidae Vjuschkov, 1957: Chroniosaurus dongusensis Tverdochlebova, 1972; Protorosauridae Huxley, 1871: Eorasaurus olsoni Sennikov, 1997; Gorgonopidae Lydekker, 1890: Sauroctonus progressus (Hartmann-Weinberg, 1938); Dicynodontidae Owen, 1859: Idelesaurus tataricus Kurkin, 2006. Material examined: Sauroctonus progressus Hartmann-Weinberg, 1938. Specimen PIN, no. 156/5, skull. Hartmann-Weinberg, 1938, pp. 51–72, text-figs. 1–5, pls. I–V (Arctognathus progressus, without no.); Tatarinov, 1974, p. 62, text-figs. 9–11, 13, and 15, pl. II, figs. 1a, 1b (lectotype); 1976, text-figs. 3, 7A, and 7B; Ivakhnenko et al., 1997, pp. 32, 74, pl. 77, figs. 1a and 1c; Ivakhnenko, 2003c, p. 368; 2005b, p. 397. Specimen PIN, no. 156/6, skull. Hartmann-Weinberg, 1938, pp. 76–80, pls. VI, VII, figs. 6, 8, and 9 (Arctognathus progressus, without no.); Tatarinov, 1966, p. 104, text-fig. 1; 1974, p. 62, pl. II, fig. 2; Ivakhnenko et al., 1997, p. 74; Ivakhnenko, 2003c, p. 368; 2005b, p. 397. Specimen 156/51, partial skull (Fig. 36a). Tatarinov, 1974, p. 62; Ivakhnenko et al., 1997, p. 74; Ivakhnenko, 2001, text-fig. 29g; 2003c, p. 368; 2005b, p. 398, textfig. 10. Specimen PIN, no. 156/56, mandible. HartmannWeinberg, 1938, pp. 75, 76 (Arctognathus progressus, without no.); Tatarinov, 1974, p. 62; Ivakhnenko et al., 1997, p. 74; Ivakhnenko, 2005b, p. 399. Specimen PIN, no. 156/60, cheek tooth (Fig. 45d). Tatarinov, 1974, p. 62; Ivakhnenko et al., 1997, p. 74; Ivakhnenko, 2005b, p. 399. Specimen PIN, no. 156/70, right angular bone. Ivakhnenko et al., 1997, p. 74; Ivakhnenko, 2005b, p. 399. Idelesaurus tataricus Kurkin, 2006. Specimen PIN, no. 156/4, skull. Kurkin, 2006, p. 83. Specimen PIN, no. 156/114, skull (Fig. 68b). Kurkin, 2006, p. 83 (holotype); Ivakhnenko, 2003c, textfig. 46. Specimen PIN, no. 156/121, left postdentary bones (Fig. 68b). Ivakhnenko, 2003b, text-fig. 4 (Oudenodon sp.). Specimen PIN, no. 156/122, left articular bone (Fig. 21e).

Specimen PIN, no. 156/126, left articular and surangular bones. Collections PIN, nos. 157; 2207 (Isheevo locality) Russian Federation, Tatarstan, Apastovskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Isheevo Faunal Assemblage, Isheevo Faunal Subassemblage. Taphonomic association: Melosauridae Fritsch, 1885: Tryphosuchus paucidens Konzhukova, 1955; Lanthanosuchidae Efremov, 1946: Lanthanosuchus watsoni Efremov, 1946; Enosuchidae Konzhukova, 1955: Enosuchus breviceps Konzhukova, 1955; Belebeyidae Ivachnenko, 2001: Permotriturus herrei Tatarinov, 1968; Syodontidae Ivachnenko, 1994: Syodon efremovi (Orlov, 1940); Anteosauridae Boonstra, 1954: Titanophoneus potens Efremov, 1938; Ulemosauridae Ivachnenko, 1994: Ulemosaurus svijagensis Riabinin, 1938; Pristerognathidae Broom, 1908: Porosteognathus efremovi Vjuschkov, 1952; Ulemicidae Ivachnenko, 1996: Ulemica invisa (Efremov, 1938). Material examined: Titanophoneus potens Efremov, 1938. Specimen PIN, no. 157/1, skeleton (Figs. 7c, 11d, 26, 39c, 48a–48g, 59a, 62b). Efremov, 1938, p. 772, text-fig. 1; 1940b, p. 39, text-fig. 11; Orlov, 1958, textfigs. 2, 3, 8a, 10, and 12 (lectotype); Tchudinov, 1983, text-fig. 15; Ivakhnenko et al., 1997, pp. 31, 63, pl. 75, fig. 2; Ivakhnenko, 2001, text-fig. 25b; 2003c, p. 364, text-fig. 13. Specimen PIN, no. 157/3, partial skeleton (Figs. 59b; 62b). Efremov, 1938, p. 772; 1940b, pl. VII, fig. 1; Orlov, 1958, text-figs. 5, 6, 9a, 9b, 11, 22b, and 30 (Doliosaurus yanschinovi, holotype); Olson, 1962, p. 66 (Doliosaurus yanschinovi); Tatarinov, 1966, p. 104 (Doliosauriscus yanschinovi); Tchudinov, 1983, textfig. 19 (Doliosauriscus yanschinovi); Ivakhnenko et al., 1997, p. 63; Ivakhnenko, 2001, text-fig. 25a; 2003c, p. 364, text-fig. 14. Specimen PIN, no. 157/186, left pterygoid (Fig. 16c). Orlov, 1958, text-fig. 23c. Specimen PIN, no. 157/221, right articular bone (Fig. 20b). Ivakhnenko et al., 1997, p. 63. Syodon efremovi (Orlov, 1940). Specimen PIN, no. 157/2, skull (Figs. 8c, 23a, 25a, 27a, 51g, 63b). Efremov, 1938, p. 772 (Titanophoneus potens, pars); 1940b, p. 43, text-fig. 12, pl. VIII, fig. 6 (Cliorhizodon sp.); Orlov, 1958, text-figs. 7a, 7b, 8a, 13a, and 31 (holotype); Tchudinov, 1983, p. 81; IvakhPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

nenko et al., 1997, pl. 72, figs. a–i; Ivakhnenko, 1995a, text-figs. 2d–2e and 3a; 2001, text-fig. 24b; 2003b, textfig. 2; 2003c, p. 264, text-fig. 12. Specimen PIN, no. 157/635, right palatine (Fig. 16b). Specimen PIN, no. 157/670, right quadrate (Fig. 20a). Orlov, 1958, text-fig. 27a. Specimen PIN, no. 157/677, mandible (Figs. 20d; 51h). Ivakhnenko et al., 1997, p. 63; Ivakhnenko, 2003c, p. 264. Specimen PIN, no. 157/1045, partial periotic (Fig. 38a). Specimen PIN, no. 157/1047, partial periotic (Fig. 34c). Ulemosaurus svijagensis Riabinin, 1938. Specimen PIN, no. 157/222, cheek tooth (Fig. 46e). Specimen PIN, no. 157/238, caniniform tooth (Fig. 46d). Specimen PIN, no. 157/243, incisor (Fig. 46c). Specimen PIN, no. 2207/1, skull: Riabinin, 1938, pls. III, V, VII, and IX (no. 1/3805); Ivakhnenko et al., 1997, p. 63; Ivakhnenko, 2003c, p. 364. Specimen PIN, no. 2207/2, partial skeleton (Figs. 14a; 62a). Riabinin, 1938, pls. IV, VI, VII, and X (no. 2/3805); Tatarinov, 1965, text-figs. 1–5 (lectotype); 1976, text-figs. 5, 11, 21, and 22; Tchudinov, 1983, text-fig. 50; Ivakhnenko et al., 1997, pl. 76, fig. 3; Ivakhnenko, 2001, text-fig. 27a; 2003c, p. 364, text-fig. 19. Specimen PIN, no. 2207/3, incisor (Fig. 46a). Riabinin, 1938, pl. XII, fig. 2 (no. 2/3805); Ivakhnenko et al., 1997, p. 63. Specimen PIN, no. 2207/14, incisor (Fig. 46b). Ulemica invisa (Efremov, 1938). Specimen PIN, no. 157/5, skull (Fig. 67b). Efremov, 1938, p. 773 (Venjukovia invisa, holotype); 1940c, pl. VIII, fig. 5 (Venjukovia invisa); Tchudinov, 1983, p. 178, text-fig. 64 (Venyukovia prima); Ivakhnenko, 1996, p. 81, text-fig. 2b (Ulemica invisa); 2001, textfig. 37b; 2003c, p. 364, text-fig. 41; Ivakhnenko et al., 1997, pl. 90, figs. a and b. Specimen PIN, no. 157/668, left quadrate-quadratojugal complex (Fig. 21a). Ivakhnenko, 1996, p. 82; Ivakhnenko et al., 1997, p. 63. Specimen PIN, no. 157/989, maxillary bone (Fig. 53d). Ivakhnenko, 1996, text-fig. 2c; Ivakhnenko et al., 1997, pl. 90, fig. d. Specimen PIN, no. 157/1112, mandible (Fig. 21b, 21c, 24a). Ivakhnenko, 1996, p. 82; 2003b, text-fig. 3; 2003c, p. 364, text-fig. 41; Ivakhnenko et al., 1997, p. 63. Specimen PIN, no. 157/1116, skull (Figs. 1b, 67b). Ivakhnenko, 1996, p. 82; 2003c, p. 364, text-fig. 41; Ivakhnenko et al., 1997, p. 63. Specimen PIN, no. 157/1117, incisor (Fig. 53c). Ivakhnenko, 1996, text-fig. 2d; Ivakhnenko et al., 1997, p. 63. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

867

Collection PIN, no. 162 (Glyadnaya Shchel’ya locality) Russian Federation, Arkhangelsk Region, Mezenskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Mezen Faunal Assemblage. Taphonomic association: Nyctiphruretidae Efremov, 1938: Nyctiphruretus acudens Efremov, 1938; Nycteroleteridae Romer, 1956: Bashkyroleter mesensis Ivachnenko, 1997; Tokosauridae Tverdochlebova et Ivachnenko, 1984: Macroleter poezicus Tverdochlebova et Ivachnenko, 1984; Varanopidae Romer et Price, 1940: Mesenosaurus romeri Efremov, 1938; Alrausuchidae fam. nov.: Alrausuchus tagax (Ivachnenko, 1990); Niaftasuchidae Ivachnenko, 1990: Niaftasuchus zekkeli Ivachnenko, 1990; Nikkasauridae Ivachnenko, 2000: Nikkasaurus tatarinovi Ivachnenko, 2000; Reiszia gubini Ivachnenko, 2000; Lepidosauria inc. sedis: Lanthanolania ivachnenkoi Modesto et Reisz, 2002. Material examined: Niaftasuchus zekkeli Ivachnenko, 1990. Specimen PIN, no. 162/63, partial skull (Fig. 11a). Ivachnenko, 2001, text-fig. 35b; 2003c, p. 364. Nikkasaurus tatarinovi Ivachnenko, 2000. Specimen PIN, no. 162/31, partial skull (Figs. 29, 30, 39a, 40f). Ivakhnenko et al., 1997, p. 59 (Therapsida ord. indet.); Ivakhnenko, 2000b, p. 184, textfigs. 1b and 1c. Specimen PIN, no. 162/33, partial skeleton (Figs. 7a, 8b, 61a). Ivakhnenko et al., 1997, p. 59 (Therapsida ord. indet.); Ivakhnenko, 2000b, p. 183, text-figs. 1a and 5a (holotype); 2001, text-figs. 41a and 41b. Reiszia gubini Ivachnenko, 2000. Specimen PIN, no. 162/32, partial skeleton (Figs. 18a, 18b, 18c, 22a, 40a–40d). Ivakhnenko et al., 1997, p. 59 (Therapsida ord. indet.); Ivakhnenko, 2000b, p. 185, text-figs. 3a, 3b, 4a, 4b, 4c, and 5b (holotype). Collection PIN, no. 164 (Belebei locality) Russian Federation, Bashkortostan, Belebeevskii District; Biarmian Series (Middle Permian), Upper Kazanian Substage; Ocher Faunal Assemblage, Ocher Faunal Subassemblage. Taphonomic association: Archegosauridae Meyer, 1858: Platyoposaurus stuckenbergi (Trautschold, 1884); Bashkirosaurus cherdyncevi Gubin, 1981; Dissorophidae Boulenger, 1902: Iratusaurus vorax Gubin, 1980;

868

IVAKHNENKO

Nycteroleteridae Romer, 1956: Bashkyroleter bashkyricus Efremov, 1940; Rhipaeosauridae Tchudinov, 1955: Rhipaeosaurus tricuspidens Efremov, 1940; Belebeyidae Ivachnenko, 2001: Belebey vegrandis Ivachnenko, 1973; Rhopalodontidae Seeley, 1894: Phthinosaurus borissiaki Efremov, 1940. Material examined: Phthinosaurus borissiaki Efremov, 1940. Specimen PIN, no. 164/7, partial mandible (Figs. 42c, 65a). Efremov, 1940a (holotype); 1954, text-fig. 55; Tatarinov, 1974, pl. VI, fig. 1; Tchudinov, 1983, text-fig. 10; Ivakhnenko, 1995b, text-figs. 1c and 1d; 2000a, text-fig. 3a; 2003c, p. 362, text-fig. 34b; Ivakhnenko et al., 1997, pl. 97, fig. 3. Collection PIN, no. 270 (Kotlovka locality) Russian Federation, Tatarstan, Elabugskii District; Biarmian Series (Middle Permian), Upper Kazanian Substage; Ocher Faunal Assemblage, Golyusherma Faunal Subassemblage. Taphonomic association: Rhopalodontidae Seeley, 1894: Rhopalodon (?) sp. Material examined: Rhopalodon (?) sp. Specimen PIN, no. 270/2, replacement canine (Fig. 42f). Ivakhnenko et al., 1997, p. 66 (Rhopalodontidae gen. indet.). An isolated replacement canine, rounded, weakly faceted in cross section. The anterior and posterior borders are well-pronounced, although serration is weak. The canine is elongated straightened, which is distinct from the canine of Microsyodon, which is similar in morphology, and is more similar to the canine of Parabradysaurus. The canine structure of Rhopalodon is not known; therefore, it is only tentatively assigned to this genus; possibly, this is a distinct primitive rhopalodontid. Collections PIN, nos. 296, 1954 (Klyuchevskii Mine-1 locality) Russian Federation, Bashkortostan, Sterlibashevskii District; Biarmian Series (Middle Permian), Upper Kazanian Substage-Urzhumian Stage; Ocher Faunal Assemblage, Ocher Faunal Subassemblage. Taphonomic association: Dissorophidae Boulenger, 1902: Zygosaurus lucius Eichwald, 1848; Phthinosuchidae Efremov, 1954: Dinosaurus murchisoni (Fischer, 1845); Rhopalodontidae Seeley, 1894: Rhopalodon wangenheimi Fischer, 1841. Material examined: Dinosaurus murchisoni (Fischer, 1845).

Specimen PIN, no. 296/1, skull: Fischer, 1845, pl. VIII, figs. a and b (Rhopalodon murchisoni, without no.); 1847, pl. VII, figs. a and b (without no.); Eichwald, 1848, pl. I, figs. a and b (Rhopalodon murchisoni, without no.); 1860, pl. LVIII, figs. 4–8 (Rhopalodon murchisoni, without no.); Efremov, 1954, pl. XIV, fig. 1 (Brithopus priscus, incorrectly designated as the holotype); Tchudinov, 1983, text-figs. 12a, 12b, and 12c (Brithopus priscus); Ivakhnenko et al., 1997, pl. 65, figs. 3a and 3b (holotype); Ivakhnenko, 2000a, p. 75; 2003c, p. 365. Specimen PIN, no. 1954/3, skull (Fig. 66a). Seeley, 1894, pl. 63 (Rhopalodon wangenheimi, without no.); Watson, 1921, p. 88, text-fig. 26 (Rhopalodon fischeri); Efremov, 1954, p. 294, text-figs. 49–53, pl. XXII, fig. I (Phthinosuchus discors, holotype); Tchudinov, 1983, text-fig. 9 (Phthinosuchus discors); Tatarinov, 1974, pp. 38–44, text-fig. 1 (Phthinosuchus discors); Ivakhnenko et al., 1997, pl. 65, fig. 3c; Ivakhnenko, 2000a, text-figs. 3c, 5; 2003c, p. 365, text-fig. 25 (incorrectly designated as no. 296/1). Collection PIN, no. 519 (Dudki locality) Russian Federation, Orenburg Region, Oktyabr’skii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Isheevo Faunal Assemblage. Taphonomic association: Ulemicidae Ivachnenko, 1996: Ulemica cf. efremovi Ivachnenko, 1995. Material examined: Ulemica cf. efremovi Ivachnenko, 1995. Specimen PIN, no. 519/1, premaxilla (Fig. 53e). Efremov, 1954, p. 221 (Venjukovia prima); Ivakhnenko, 1996, p. 82; 2003c, p. 363; Ivakhnenko et al., 1997, p. 61. Collection PIN, no. 520 (Malyi Uran locality) Russian Federation, Orenburg Region, Krasnogvardeiskii District; Biarmian Series (Middle Permian), Urzhumian Stage; Isheevo Faunal Assemblage, Isheevo Subassemblage. Taphonomic association: Melosauridae Fritsch, 1885: Konzhukovia vetusta (Konzhukova, 1955); Lanthanosuchidae Efremov, 1946: Chalcosaurus sp.; Syodontidae Ivachnenko, 1994: Syodon sp.; Anteosauridae Boonstra, 1954: Titanophoneus adamanteus (Orlov, 1958); Pristerognathidae Broom, 1908: Porosteognathus (?) sp. Material examined: Titanophoneus adamanteus (Orlov, 1958). PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

Specimen PIN, no. 520/30, skull. Orlov, 1958, textfig. 32a (Doliosaurus adamanteus, holotype, incorrectly designated as no. 519/1); Ivakhnenko et al., 1997, p. 31, pl. 76, fig. 1; Ivakhnenko, 2003c, p. 366, text-fig. 15. Collection PIN, no. 1536 (Berezhane locality) Russian Federation, Kirov Region, Slobodskoi District; Tatarian Series (Upper Permian), Upper Vyatkian Substage, Vyatkian Horizon; Sokolki Faunal Assemblage, Sokolki Faunal Subassemblage. Taphonomic association: Chroniosuchidae gen. indet.; Dicynodontidae Owen, 1859: Vivaxosaurus permirus Kalandadze et Kurkin, 2000; Delectosaurus berezhanensis Kurkin, 2001. Material examined: Delectosaurus berezhanensis Kurkin, 2001. Specimen PIN, no. 1536/2, skull (Fig. 37b). Ivakhnenko et al., 1997, p. 56. (Dicynodon sp., determ.); Kurkin, 2001, p. 57, text-fig. 3 (holotype); Ivakhnenko, 2003c, text-fig. 51. Specimen PIN, no. 1536/25, perioticum. Vivaxosaurus permirus Kalandadze et Kurkin, 2000. Specimen PIN, no. 1536/1, skull (Fig. 21d). Ivakhnenko et al., 1997, p. 56. (Dicynodon sp., determ.); Kalandadze and Kurkin, 2000, text-figs. 1–3 (holotype); Ivakhnenko, 2003c, text-fig. 49. Collection PIN, no. 1538 (Purly locality) Russian Federation, Nizhni Novgorod Region, Tonshaevskii District; Tatarian Series (Upper Permian), Upper Vyatkian Substage, Vyatkian Horizon; Vyazniki Faunal Assemblage. Taphonomic association: Dvinosauridae Amalitzky, 1921: Dvinosaurus purlensis Shishkin, 1968; Chroniosuchidae Vjuschkov, 1957: Uralerpeton tverdochlebovae Golubev, 1998; Proterosuchidae Huene, 1908: Archosaurus rossicus Tatarinov, 1960; Dicynodontidae gen. indet.; Scaloposauridae Haughton, 1924: Malasaurus germanus Tatarinov, 2002; Whaitsiidae gen. indet.; Nanictidopidae Watson et Romer, 1956: Hexacynodon purlinensis Tatarinov, 1974. Material examined: Dicynodontidae gen. indet. Specimen PIN, no. 1538/60, replacement canine (Fig. 55c). Specimen PIN, no. 1538/61, canine (Fig. 55d). PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

869

Collection PIN, no. 1758 (Ezhovo locality) Russian Federation, Permian Region, Ocherskii District; Biarmian Series (Middle Permian), Upper Kazanian Substage; Ocher Faunal Assemblage, Ocher Faunal Subassemblage. Taphonomic association: Archegosauridae Meyer, 1858: Collidosuchus tchudinovi Gubin, 1986; Melosauridae Fritsch, 1885: Konzhukovia tarda Gubin, 1991; Dissorophidae Boulenger, 1902: Kamacops acervalis Gubin, 1980; Eotitanosuchidae Tchudinov, 1960: Biarmosuchus tener Tchudinov, 1960; Archaeosyodontidae Ivachnenko, fam. nov.: Archaeosyodon praeventor Tchudinov, 1960; Estemmenosuchidae Tchudinov, 1960: Estemmenosuchus uralensis Tchudinov, 1960; Estemmenosuchus mirabilis Tchudinov, 1968; Venyukoviidae Efremov, 1940: Otsheria netzvetajevi Tchudinov, 1960. Material examined: Biarmosuchus tener Tchudinov, 1960. Specimen PIN, no. 1758/1, skull (Figs. 58c, 64b). Tchudinov, 1960, p. 82, text-fig. 1 (Eotitanosuchus olsoni, holotype); 1983, pl. 1-I, fig. 2 (Eotitanosuchus olsoni); Chudinov, 1965, text-fig. 2 (Eotitanosuchus olsoni); Sigogneau and Tchudinov, 1972, text-figs. 24– 26 (Eotitanosuchus olsoni); Ivakhnenko et al., 1997, pl. 65, fig. 1d; Ivakhnenko, 1999, text-fig. 1d; 2001, text-fig. 23a; 2003c, p. 364, text-fig. 5b. Specimen PIN, no. 1758/2, skeleton (Figs. 12a, 12b, 58a). Tchudinov, 1960, p. 84, text-fig. 2 (holotype); 1983, text-figs. 3a and 3b; Sigogneau and Tchudinov, 1972, text-figs. 1–3, 5, 7–10; Ivakhnenko et al., 1997, pl. 65, figs. 1a–1c; Ivakhnenko, 1999, text-fig. 1a; 2001, text-fig. 23b; 2003c, p. 364, text-fig. 5a. Specimen PIN, no. 1758/7, skeleton. Tchudinov, 1964b, text-fig. 1 (Biarmosaurus antecessor, holotype); 1983, text-fig. 4; Sigogneau and Tchudinov, 1972, text-figs. 15–19 (Biarmosuchus cf. tener); Ivakhnenko, 1999, p. 290; 2003c, p. 364; Ivakhnenko et al., 1997, p. 61. Specimen PIN, no. 1758/8, skull (Fig. 58b). Tchudinov, 1964b, text-fig. 2; 1983, p. 57; Sigogneau and Tchudinov, 1972, text-figs. 11–14; Ivakhnenko et al., 1997, p. 61; Ivakhnenko, 1999, text-fig. 1b; 2003c, p. 364. Specimen PIN, no. 1758/18, skull (Fig. 16a). Tchudinov, 1964b, p. 90; 1983, p. 57; Sigogneau and Tchudinov, 1972, text-fig. 27; Ivakhnenko et al., 1997, p. 61; Ivakhnenko, 1999, text-fig. 1c; 2003c, p. 364. Specimen PIN, no. 1758/19, skull (Figs. 12a, 12b, 32d). Tchudinov, 1964b, p. 90; 1983, p. 48; Sigogneau and Tchudinov, 1972, text-fig. 4; Ivakhnenko et al., 1997, p. 61; Ivakhnenko, 1999, p. 290; 2003c, p. 364.

870

IVAKHNENKO

Specimen PIN, no. 1758/85, skull (Figs. 10a, 12a, 12b, 32a–32c, 33). Sigogneau and Tchudinov, 1972, text-fig. 27; Tchudinov, 1983, p. 66; Ivakhnenko et al., 1997, p. 61; Ivakhnenko, 1999, p. 290; 2003c, p. 364. Specimen PIN, no. 1758/86, skeleton (Figs. 19a– 19d, 22c, 58b, 64b). Sigogneau and Tchudinov, 1972, text-fig. 20; Tchudinov, 1983, p. 57; Ivakhnenko et al., 1997, p. 61; Ivakhnenko, 1999, p. 290; 2003c, p. 364. Specimen PIN, no. 1758/212, canine (Fig. 50d). Ivakhnenko et al., 1997, pl. 65, fig. 1e. Specimen PIN, no. 1758/255, skull (Figs. 16a, 17c, 50e, 50f). Tchudinov, 1983, p. 57; Ivakhnenko et al., 1997, p. 61; Ivakhnenko, 1999, p. 290; 2003c, p. 364. Specimen PIN, no. 1758/292, right maxillary. Sigogneau and Tchudinov, 1972, text-fig. 11 (Eotitanosuchus olsoni); Tchudinov, 1983, text-fig. 8 (Ivantosaurus ensifer, holotype); Ivakhnenko et al., 1997, p. 61; Ivakhnenko, 1999, p. 290; 2003c, p. 364. Specimen PIN, no. 1758/307, dentary (Figs. 22c, 64b). Specimen PIN, no. 1758/334, incisor (Fig. 50c). Estemmenosuchus uralensis Tchudinov, 1960. Specimen PIN, no. 1758/4, skull. Tchudinov, 1960, text-fig. 4 (holotype); 1983, text-fig. 4; Ivakhnenko et al., 1997, p. 29; Ivakhnenko, 2000a, text-fig. 2a; 2003c, p. 364. Specimen PIN, no. 1758/22, skeleton (Figs. 23b, 60d). Tchudinov, 1983, p. 124; Ivakhnenko et al., 1997, p. 61; Ivakhnenko, 2000a, text-fig. 2b, 3b; 2001, textfig. 29b; 2003c, p. 364. Specimen PIN, no. 1758/23, skeleton. Tchudinov, 1983, p. 124; Ivakhnenko et al., 1997, p. 61; Ivakhnenko, 2000a, p. 73; 2003c, p. 364. Specimen PIN, no. 1758/27, skull. Tchudinov, 1983, text-fig. 4; Ivakhnenko et al., 1997, p. 61; Ivakhnenko, 2000a, p. 73; 2003c, p. 364. Specimen PIN, no. 1758/79, skeleton (Fig. 60b). Tchudinov, 1968, text-fig. 2 (Anoplosuchus tenuirostris, holotype); 1983, text-fig. 38; Ivakhnenko et al., 1997, p. 61; Ivakhnenko, 2000a, text-fig. 4b; 2003c, p. 364, text-fig. 37b. Specimen PIN, no. 1758/80, partial calvarium (Fig. 2b). Tchudinov, 1968, text-fig. 2 (Anoplosuchus tenuirostris); 1983, text-figs. 39a, 39b; Ivakhnenko et al., 1997, p. 61; Ivakhnenko, 2000a, p. 73. Specimen PIN, no. 1758/82, partial calvarium. Tchudinov, 1968, p. 30 (Anoplosuchus tenuirostris); 1983, p. 139 (Anoplosuchus tenuirostris); Ivakhnenko, 2000a, p. 73; 2003c, p. 364. Specimen PIN, no. 1758/200, partial calvarium (Fig. 23b). Specimen PIN, no. 1758/227, mandible (Fig. 23b). Tchudinov, 1968, text-fig. 4 (Anoplosuchus tenuirostris); 1983, p. 139 (Anoplosuchus tenuirostris); Ivakhnenko et al., 1997, p. 61; Ivakhnenko, 2000a, p. 73; 2003b, text-fig. 1.

Specimen PIN, no. 1758/300, partial skeleton (Fig. 60a). Tchudinov, 1983, text-fig. 44 (Zopherosuchus luceus, holotype); Ivakhnenko, 2000a, text-fig. 4a; 2003c, p. 364, text-fig. 37a. Specimen PIN, no. 1758/325, right parietal (Fig. 2a). Tchudinov, 1983, p. 124; Ivakhnenko et al., 1997, p. 61. Specimen PIN, no. 1758/331, partial calvarium (Fig. 60c). Tchudinov, 1983, p. 124; Ivakhnenko et al., 1997, p. 61; Ivakhnenko, 2000a, p. 73. Estemmenosuchus mirabilis Tchudinov, 1968. Specimen PIN, no. 1758/6, skeleton (Figs. 4b, 15a, 27b, 43e, 65c). Chudinov, 1965, text-figs. 3a and 3b (Estemmenosuchus uralensis); Tchudinov, 1968, textfig. 1 (holotype); 1983, text-figs. 35a–35e; Ivakhnenko et al., 1997, pl. 69; Ivakhnenko, 2000a, text-fig. 1; 2001, text-fig. 29d; 2003c, p. 364, text-fig. 38. Specimen PIN, no. 1758/336, cheek tooth (Fig. 40f). Archaeosyodon praeventor Tchudinov, 1960. Specimen PIN, no. 1758/3, skull. Tchudinov, 1960, text-fig. 3 (holotype); 1983, text-fig. 16; Ivakhnenko et al., 1997, pl. 71, fig. 2a; Ivakhnenko, 2003c, p. 364. Specimen PIN, no. 1758/93, partial skull (Figs. 11c, 15b). Tchudinov, 1983, p. 85; Ivakhnenko et al., 1997, p. 62; Ivakhnenko, 2003c, p. 364. Specimen PIN, no. 1758/95, skull (Figs. 51c, 51f). Tchudinov, 1983, pl. III, fig. 1; Ivakhnenko et al., 1997, pl. 71, fig. 2c. Specimen PIN, no. 1758/118, left maxillary (Fig. 10b). Specimen PIN, no. 1758/293, skull (Fig. 63a). Tchudinov, 1983, p. 89; Ivakhnenko et al., 1997, p. 62; Ivakhnenko, 2003c, p. 364. Specimen PIN, no. 1758/297, partial skull (Fig. 63a). Tchudinov, 1983, p. 89; Ivakhnenko et al., 1997, p. 62. Specimen PIN, no. 1758/315, canine (Fig. 51d) Tchudinov, 1983, p. 89; Ivakhnenko et al., 1997, p. 62. Specimen PIN, no. 1758/328, right dentary (Fig. 51e). Ivakhnenko et al., 1997, p. 112, pl. 2, fig. e. Otsheria netzvetajevi Tchudinov, 1960. Specimen PIN, no. 1758/5, skull (Fig. 67a). Tchudinov, 1960, text-fig. 5 (holotype); 1983, text-figs. 61a and 61b; Chudinov, 1965, text-fig. 4; Ivakhnenko, 1996, text-figs. 1a and 1b; 2001, text-fig. 36a; 2003c, p. 364, text-fig. 40; Ivakhnenko et al., 1997, pl. 89, figs. 2a and 2b. Collection PIN, no. 1954 (Staro-Myasnikovskii Mine locality) Russian Federation, Orenburg Region, Oktyabr’skii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Isheevo Faunal Assemblage, Malaya Kinel Faunal Subassemblage. Taphonomic association: Deuterosauridae Seeley, 1894: Deuterosaurus jubilaei (Nopcsa, 1928). Material examined: PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

Deuterosaurus jubilaei (Nopcsa, 1928). Specimen PIN, no. 1954/2, skull. Seeley, 1894, p. 680 (Deuterosaurus biarmicus, partim); Nopcsa, 1928, p. 12, pl. 3 (Mnemejosaurus jubilaei, holotype); Ivakhnenko, 2003c, p. 368 (holotype); Nopcsa, 1928, p. 14, pl. 4 (Uraniscosaurus watsoni, holotype); Efremov, 1954, p. 193–195 (Deuterosaurus biarmicus, partim); Tchudinov, 1983, text-fig. 23 (Mnemejosaurus jubilaei); Ivakhnenko et al., 1997, pp. 31, 75, pl. 74A; Ivakhnenko, 2003c, p. 387, text-fig. 18. Collection PIN, nos. 1954, 1955 (Klyuchevskii Mine-2 locality) Russian Federation, Bashkortostan, Sterlibashevskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Isheevo Faunal Assemblage, Malaya Kinel Faunal Subassemblage. Taphonomic association: Deuterosauridae Seeley, 1894: Deuterosaurus biarmicus Eichwald, 1846; Ulemosauridae Ivachnenko, 1994: Ulemosaurus gigas (Efremov, 1954). Material examined: Deuterosaurus biarmicus Eichwald, 1846. Specimen PIN, no. 1954/1, partial skull (Figs. 17b, 47a–47d, 62c). Eichwald, 1860, pl. LVIII, figs. 1 and 2; Efremov, 1954, pls. I, II, III (“holotype”); Tchudinov, 1983, text-fig. 26; Ivakhnenko et al., 1997, pl. 73, fig. 3; Ivakhnenko, 2003c, p. 365, text-fig. 17. Ulemosaurus gigas (Efremov, 1954). Specimen PIN, no. 1955/3, incisor. Efremov, 1954, text-fig. 21. Specimen PIN, no. 1955/5, incisor. Efremov, 1954, text-fig. 20; Ivakhnenko et al., 1997, p. 32, pl. 76, fig. 2 (lectotype, incorrectly designated as no. 1955/3); Ivakhnenko, 2003c, p. 365, text-figs. 20e and 20f. Collection PIN, no. 1954 (unknown locality) Russian Federation, Orenburg Region, Kargala River mines; Biarmian Series (Middle Permian); Isheevo Faunal Assemblage. Taphonomic association: Phthinosuchidae Efremov, 1954: Admetophoneus kargalensis Efremov, 1954. Material examined: Admetophoneus kargalensis Efremov, 1954. Specimen PIN, no. 1954/5, partial right maxillary. Efremov, 1954, p. 260, text-fig. 45 (one of “holotypes”); Tchudinov, 1983, p. 99, text-fig. 24 (lectotype); Ivakhnenko et al., 1997, p. 39, pl. 97, fig. 3 (Titanophoneus sp.); Ivakhnenko, 2003c, p. 386 (Titanophoneus sp.). This form was determined as a “brithopodid” (Efremov, 1954; Tchudinov, 1983) or Titanophoneus sp. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

871

However, the study of the material of Phthinosuchidae has shown that the structure of the canine and postcanines is characteristic of this group. In addition, Admetophoneus has a sheath for the replacement canine, which has not been found in Anteosauridae. Collection PIN, no. 1955 (Butlerovka locality) Russian Federation, Tatarstan, Alekseevskii District; Biarmian Series (Middle Permian), Urzhumian Stage; Isheevo Faunal Assemblage, Isheevo Faunal Subassemblage. Taphonomic association: Anteosauridae Boonstra, 1954: Titanophoneus rugosus (Trautschold, 1884). Material examined: Titanophoneus rugosus (Trautschold, 1884). Specimen PIN, no. 1955/1a, angular bone. Trautschold, 1884, pl. VIII, fig. 1a (without no.); Efremov, 1954, p. 266 (Titanophoneus sp.); Ivakhnenko et al., 1997, text-fig. 1 (lectotype); Ivakhnenko, 2003c, p. 363, text-fig. 16. Collection PIN, no. 2005 (Sokolki locality) Russian Federation, Arkhangelsk Region, Kotlasskii District; Tatarian Series (Upper Permian), Upper Vyatkian Substage, Vyatkian Horizon; Sokolki Faunal Assemblage, Sokolki Faunal Subassemblage. Taphonomic association: Dvinosauridae Amalitzky, 1921: Dvinosaurus primus Amalitzky, 1921; Karpinskiosauridae Sushkin, 1925: Karpinskiosaurus secundus (Amalitzky, 1921); Kotlassiidae Romer, 1934: Kotlassia prima Amalitzky, 1921; Pareiasauridae Seeley, 1888: Scutosaurus karpinskii (Amalitzky, 1922); Scutosaurus tuberculatus (Amalitzky, 1922); Chroniosuchidae Vjuschkov, 1957: Chroniosuchus licharevi (Riabinin, 1962); Gorgonopidae Lydekker, 1890: Pravoslavlevia parva (Pravoslavlev, 1927); Inostranceviidae Huene, 1948: Inostrancevia latifrons Pravoslavlev, 1927; Inostrancevia alexandri Amalitzky, 1922; Dicynodontidae Owen, 1859: Dicynodon trautscholdi Amalitzky, 1922; Dicynodon amalitzkii Sushkin, 1922; Elph borealis Kurkin, 1999; Annatherapsididae Kuhn, 1963: Annatherapsidus petri (Amalitzky, 1922); Dviniidae Sushkin, 1928: Dvinia prima Amalitzky, 1922. Material examined: Inostrancevia latifrons Pravoslavlev, 1927.

872

IVAKHNENKO

Specimen PIN, no. 2005/1715, left maxillary (Fig. 57b). Specimen PIN, no. 2005/1857, skull. Pravoslavlev, 1927a, text-figs. 33–37 (holotype, without no.); Tatarinov, 1974, pl. V, fig. 2; Ivakhnenko et al., 1997, pl. 79, fig. 1; Ivakhnenko, 2003c, p. 368, text-fig. 30. Inostrancevia alexandri Amalitzky, 1922. Specimen PIN, no. 2005/1587, skeleton (Fig. 66c). Amalitzky, 1922, text-fig. 9 (without no.); Pravoslavlev, 1927a, pls. I, II, III, figs. 1–5, 7, and 8; Tatarinov, 1974, pl. 4, text-fig. 1 (lectotype); Ivakhnenko et al., 1997, pl. 78; Ivakhnenko, 2001, text-fig. 34; 2003c, p. 368, text-fig. 29. Specimen PIN, no. 2005/1588, skeleton. Amalitzky, 1922, p. 8 (without no.); Pravoslavlev, 1927a, pl. V, figs. 11–14; VI, VII (without no.); Tatarinov, 1974, p. 89; Ivakhnenko et al., 1997, p. 75; Ivakhnenko, 2003c, p. 368. Specimen PIN, no. 2005/2099, partial calvarium. Pravoslavlev, 1927b, text-fig. 1 (without no.); Tatarinov, 1974, pl. IV, fig. 2; Ivakhnenko, 2000a, text-fig. 6; 2003c, p. 368. Inostrancevia sp. Specimen PIN, no. 2005/1707, angular bone (Fig. 25b). Specimen PIN, no. 2005/1716, left maxillary. Specimen PIN, no. 2005/1757, upper canine (Fig. 44c). Specimen PIN, no. 2005/1775, incisor (Fig. 44b). Specimen PIN, no. 2005/2265, left articular bone (Fig. 20c). Ivakhnenko, 2003b, p. 290. Pravoslavlevia parva (Pravoslavlev, 1927). Specimen PIN, no. 2005/1859, skull. Pravoslavlev, 1927a, pp. 70–75, pl. VIII, figs. 29–32 (Inostrancevia parva, holotype); Vjuschkov, 1953, p. 399; Tatarinov, 1974, pp. 86–88, pl. III, text-fig. 2; Ivakhnenko et al., 1997, pp. 32, 75, pl. 79, fig. 2; Ivakhnenko, 2002b, pp. 56–65, text-figs. 2a–2c; Ivakhnenko, 2003c, pp. 368, 396, text-fig. 22; Ivakhnenko, 2005b, p. 400. Dicynodon trautscholdi Amalitzky, 1922. Specimen PIN, no. 2005/1, skull. Amalitzky, 1922, p. 4 (holotype); Ivakhnenko, 2003c, text-fig. 47. Dicynodon amalitzkii Sushkin, 1922. Specimen PIN, no. 2005/38, skull (Fig. 68c). Sushkin, 1922, p. 9 (holotype); Ivakhnenko, 2003c, text-fig. 48. Collection PIN, no. 2212 (Kotelnich locality) Russian Federation, Kirov Region, Kotel’nichskii District; Tatarian Series (Upper Permian), Upper Severodvinian Substage, Severodvinian Horizon; Sokolki Faunal Assemblage, Kotelnich Faunal Subassemblage. Taphonomic association: Nycteroleteridae Romer, 1956: Emeroleter levis Ivachnenko, 1997;

Bradysauridae Huene, 1948: Deltavjatia vjatkensis (Hartmann-Weinberg, 1937); Scylacosauridae Broom, 1903: Kotelcephalon viatkensis Tatarinov, 1999; Chthonosauridae Tatarinov, 1974: Viatkosuchus sumini Tatarinov, 1995; Phthinosuchidae Efremov, 1954: Viatkogorgon ivakhnenkoi Tatarinov, 1999; Scaloposauridae Haughton, 1924: Scalopodon tenuisfrons Tatarinov, 1999; Scalopodontes kotelnichi Tatarinov, 2000; Perplexisauridae Tatarinov, 2000: Perplexisaurus foveatus Tatarinov, 1997; Chlynovia serridentatus Tatarinov, 2000; Karenitidae Tatarinov, 1997: Karenites ornamentatus Tatarinov, 1995; Galeopidae Broom, 1912: Suminia getmanovi Ivachnenko, 1994. Material examined: Viatkogorgon ivakhnenkoi Tatarinov, 1999. Specimen PIN, no. 2212/61, skeleton (Figs. 7b, 8a). Ivakhnenko et al., 1997, p. 66 (Gorgonopidae gen. indet.); Tatarinov, 1999a, text-fig. 1 (holotype); Ivakhnenko, 2003c, p. 365, text-figs. 27a and 27c. Suminia getmanovi Ivachnenko, 1994. Specimen PIN, no. 2212/10, skeleton (Figs. 13a, 13b, 67c). Ivakhnenko, 1994, text-fig. 1 (holotype); 2001, text-fig. 38b; 2003c, p. 365; Ivakhnenko et al., 1997, pl. 92, figs. A and B. Specimen PIN, no. 2212/18, premaxillary and maxillary bones (Fig. 57a). Ivakhnenko, 1994, p. 77; Ivakhnenko et al., 1997, p. 66; Rybczynski, 2000, p. 332. Specimen PIN, no. 2212/31, skull (Fig. 56c). Ivakhnenko, 1994, p. 77; 2003c, p. 365; Ivakhnenko et al., 1997, p. 66. Specimen PIN, no. 2212/32, skull. Ivakhnenko, 1994, text-fig. 2b; 2001, text-fig. 38a; 2003c, p. 365; Ivakhnenko et al., 1997, p. 66. Specimen PIN, no. 2212/33, cheek tooth (Figs. 54c, 54d). Specimen PIN, no. 2212/62, skull (Figs. 1a, 13a, 13b, 17a, 24b, 37a, 67c). Ivakhnenko et al., 1997, pl. 92, figs. c and d; Rybczynski, 2000, text-figs. 6–13; Ivakhnenko, 2003c, p. 365, text-fig. 43. Specimen PIN, no. 2212/82, dentaries (Figs. 54a, 54b). Ivakhnenko, 1996, text-fig. 2f; Ivakhnenko et al., 1997, pl. 92, fig. 7. Specimen PIN, no. 2212/99, dentaries (Fig. 56b). Ivakhnenko et al., 1997, p. 66. Collection PIN, nos. 2353, 2356 (Zavrazh’e locality) Russian Federation, Arkhangelsk Region, Kotlasskii District; Tatarian Series (Upper Permian), Upper Vyatkian Substage, Vyatkian Horizon; Sokolki Faunal Assemblage, Sokolki Faunal Subassemblage. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

Taphonomic association: Dvinosauridae Amalitzky, 1921: Dvinosaurus primus Amalitzky, 1921; Pareiasauridae Seeley, 1888: Scutosaurus karpinskii (Amalitzky, 1922); Chroniosuchidae Vjuschkov, 1957: Chroniosuchus licharevi (Riabinin, 1962); Inostranceviidae Huene, 1948: Inostrancevia latifrons Pravoslavlev, 1927; Dicynodontidae Owen, 1859: Dicynodon trautscholdi Amalitzky, 1922; Elph borealis Kurkin, 1999. Material examined: Inostrancevia latifrons Pravoslavlev, 1927. Specimen PIN, no. 2356/32, skull. Vjuschkov, 1953, p. 399; Tatarinov, 1974, p. 93; Ivakhnenko et al., 1997, p. 62; Ivakhnenko, 2003c, p. 370. Elph borealis Kurkin, 1999. Specimen PIN, no. 2353/37, skull. Kurkin, 1999, text-figs. 1 and 2 (holotype); Ivakhnenko, 2003c, textfig. 52. Collection PIN, no. 2416 (Agafonovo locality) Russian Federation, Kirov Region, Kotelnichskii District; Tatarian Series (Upper Permian), Upper Severodvinian Substage, Severodvinian Horizon; Sokolki Faunal Assemblage, Ilinskoe Faunal Subassemblage. Taphonomic association: Dvinosauridae Amalitzky, 1921: Dvinosaurus primus Amalitzky, 1921; Pareiasauridae Seeley, 1888: Proelginia sp. Chroniosuchidae Vjuschkov, 1957: Chroniosaurus levis Golubev, 1998; Burnetiidae Broom, 1923: Proburnetia viatkensis Tatarinov, 1968. Material examined: Proburnetia viatkensis Tatarinov, 1968. Specimen PIN, no. 2416/1, skull (Figs. 4a, 28b, 65b). Tatarinov, 1968, text-fig. 2 (holotype); 1974, textfigs. 4–8; Sigogneau-Russell, 1989, text-figs. 50–52; Ivakhnenko et al., 1997, pl. 66, fig. 1; Ivakhnenko, 2000a, text-fig. 7; 2001, text-figs. 32a and 32b; 2003c, text-fig. 32. Collection PIN, no. 2505 (Zhaksy-Kargala locality) Kazakhstan, Aktyubinsk Region; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Isheevo Faunal Assemblage, Malaya Kinel Faunal Subassemblage. Taphonomic association: Syodontinae Ivachnenko, 1994: Syodon gusevi (Tchudinov, 1968). Material examined: PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

873

Syodon gusevi (Tchudinov, 1968). Specimen PIN, no. 2505/1, partial skull (Figs. 34b, 35a, 35b). Tchudinov, 1968, p. 5, text-figs. 1–3 (Notosyodon gusevi, holotype); 1983, p. 91, text-figs. 19–21 (Notosyodon gusevi); Ivakhnenko, 1995a, text-figs. 2c and 3b (“Notosyodon gusevi”); 2003c, p. 370, textfig. 11; Ivakhnenko et al., 1997, p. 30, pl. 73, text-fig. 1. Collection PIN, no. 2629 (Nezhinka locality) Russian Federation, Orenburg Region, Orenburgskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Isheevo Faunal Assemblage, Malaya Kinel Faunal Subassemblage. Taphonomic association: Deuterosauridae Seeley, 1894: Deuterosaurus jubilaei (Nopcsa, 1928). Material examined: Deuterosaurus jubilaei (Nopcsa, 1928). Specimen PIN, no. 2629/1, skull (Fig. 62c). Tchudinov, 1983, pl. IV, fig. 1 (Mnemeiosaurus jubilaei); Ivakhnenko et al., 1997, pl. 74, fig. B; Ivakhnenko, 2001, text-fig. 26a; 2003c, p. 366, text-fig. 18. Collection PIN, no. 2793 (Novo-Nikol’skoe-3 locality) Russian Federation, Orenburg Region, Aleksandrovskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Isheevo Faunal Assemblage, Isheevo Faunal Subassemblage. Taphonomic association: Melosauridae Fritsch, 1885: Konzhukovia vetusta (Konzhukova, 1955); Lanthanosuchidae Efremov, 1946: Chalcosaurus lukjanovae (Ivachnenko, 1980); Anteosauridae Boonstra, 1954: Titanophoneus sp.; Ulemicidae Ivachnenko, 1996: Ulemica efremovi Ivachnenko, 1995; Procynosuchidae gen. indet. Material examined: Ulemica efremovi Ivachnenko, 1995. Specimen PIN, no. 2793/1, skull. Tchudinov, 1983, text-fig. 63 (Venyukovia prima, pars); Ivakhnenko, 1996, text-fig. 3 (holotype); 2001, text-fig. 37b; 2003c, text-fig. 42; Ivakhnenko et al., 1997, pl. 91. Collection PIN, no. 2896 (Blumental-3 locality) Russian Federation, Orenburg Region, Belyaevskii District; Tatarian Series (Upper Permian), Upper Vyatkian Substage, Vyatkian Horizon; Sokolki Faunal Assemblage, Sokolki Faunal Subassemblage. Taphonomic association: Dvinosauridae Amalitzky, 1921: Dvinosaurus sp.; Karpinskiosauridae Sushkin, 1925: Karpinskiosaurus sp.;

874

IVAKHNENKO

Pareiasauridae Seeley, 1888: Scutosaurus sp.; Chroniosuchidae Vjuschkov, 1957: Chroniosuchus paradoxus Vjuschkov, 1957; Inostranceviidae Huene, 1948: Inostrancevia uralensis Tatarinov, 1974; Annatherapsididae Kuhn, 1963: Annatherapsidus cf. petri (Amalitzky, 1922); Procynosuchidae Broom, 1937: Uralocynodon tverdokhlebovae Tatarinov, 1987. Material examined: Inostrancevia uralensis Tatarinov, 1974. Specimen PIN, no. 2896/1, periotic (Fig. 36b). Tatarinov, 1974, text-figs. 17 and 18 (holotype); Sigogneau-Russell, 1989, text-fig. 271; Ivakhnenko et al., 1997, pl. 77, fig. 2; Ivakhnenko, 2003c, p. 363. Specimen PIN, no. 2896/2, periotic. Tatarinov, 1974, text-figs. 19 and 20; Ivakhnenko et al., 1997, p. 56 (incorrectly designated as no. 2896/3); Ivakhnenko, 2003c, p. 363. Collection PIN, no. 3159 (Navoloki locality) Russian Federation, Vologda Region, Niuksenitskii District; Tatarian Series (Upper Permian), Upper Severodvinian Substage, Severodvinian Horizon; Sokolki Faunal Assemblage, Ilinskoe Faunal Subassemblage. Taphonomic association: Dvinosauridae Amalitzky, 1921: Dvinosaurus sp.; Kotlassiidae Romer, 1934: Microphon cf. exiguus Ivachnenko, 1983; Chroniosuchidae Vjuschkov, 1957: Chroniosaurus dongusensis Tverdochlebova, 1972; Burnetiidae Broom, 1923: Niuksenitia sukhonensis Tatarinov, 1977; Gorgonopia fam. indet.; Galeopidae Broom, 1912: Suminia cf. getmanovi Ivachnenko, 1994. Material examined: Niuksenitia sukhonensis Tatarinov, 1977. Specimen PIN, no. 3159/1, partial skull. Tatarinov, 1977, text-figs. 1 and 2 (holotype); Sigogneau-Russell, 1989, text-figs. 255 and 256; Ivakhnenko et al., 1997, pl. 66, fig. 2; Ivakhnenko, 2002b, text-fig. 1; 2003c, p. 365, text-fig. 33. Collection PIN, no. 3586 (Ust’-Peza locality) Russian Federation, Arkhangelsk Region, Mezenskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Mezen Faunal Assemblage. Taphonomic association: Nyctiphruretidae Efremov, 1938: Nyctiphruretus acudens Efremov, 1938;

Nycteroleteridae Romer, 1956: Bashkyroleter mesensis Ivachnenko, 1997; Tokosauridae Tverdochlebova et Ivachnenko, 1984: Macroleter poezicus Tverdochlebova et Ivachnenko, 1984; Varanopidae Romer et Price, 1940: Mesenosaurus romeri Efremov, 1938; Alrausuchidae fam. nov.: Alrausuchus tagax (Ivachnenko, 1990). Material examined: Alrausuchus tagax (Ivachnenko, 1990). Specimen PIN, no. 3586/6, partial skull (Fig. 8d). Ivakhnenko, 1990, p. 87 (Biarmosuchus tagax); 1999, p. 80 (Biarmosuchus tagax); 2003c, p. 369 (Biarmosuchus tagax); Ivakhnenko et al., 1997, p. 78 (Biarmosuchus tagax). Specimen PIN, no. 3586/14, partial dentary (Figs. 22b, 64a). Ivakhnenko, 1990, p. 87 (Biarmosuchus tagax); 1999, p. 80 (Biarmosuchus tagax); 2003c, p. 369 (Biarmosuchus tagax); Ivakhnenko et al., 1997, p. 78 (Biarmosuchus tagax). Collection PIN, no. 3706 (Peza-1 locality) Russian Federation, Arkhangelsk Region, Mezenskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Mezen Faunal Assemblage. Taphonomic association: Nyctiphruretidae Efremov, 1938: Nyctiphruretus acudens Efremov, 1938; Nycteroleteridae Romer, 1956: Nycteroleter ineptus Efremov, 1938; Bashkyroleter mesensis Ivachnenko, 1997; Tokosauridae Tverdochlebova et Ivachnenko, 1984: Macroleter poezicus Tverdochlebova et Ivachnenko, 1984; Lanthaniscidae Ivachnenko, 2007: Lanthaniscus efremovi Ivachnenko, 1980; Varanopidae Romer et Price, 1940: Mesenosaurus romeri Efremov, 1938; Alrausuchidae fam. nov.: Alrausuchus tagax (Ivachnenko, 1990); Niaftasuchidae Ivachnenko, 1990: Niaftasuchus zekkeli Ivachnenko, 1990. Material examined: Alrausuchus tagax (Ivachnenko, 1990). Specimen PIN, no. 3706/10, partial skull (Figs. 3b, 7d, 11b, 22b, 28a, 31, 34a, 39b, 64a). Ivakhnenko, 1990, text-fig. 2 (Biarmosuchus tagax, holotype); 1999, text-fig. 4 (Biarmosuchus tagax); 2002a, text-fig. 2; 2003b, p. 289 (Biarmosuchus tagax); 2003c, text-fig. 6 (Biarmosuchus tagax); Ivakhnenko et al., 1997, pl. 65, fig. 2. (Biarmosuchus tagax). Specimen PIN, no. 3706/17, left dentary (Fig. 22b). Ivakhnenko et al., 1997, p. 71; (Therapsida, gen. nov.). PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

Collection PIN, no. 3717 (Ust’-Nyafta locality) Russian Federation, Arkhangelsk Region, Mezenskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Mezen Faunal Assemblage. Taphonomic association: Nyctiphruretidae Efremov, 1938: Nyctiphruretus acudens Efremov, 1938; Nycteroleteridae Romer, 1956: Bashkyroleter mesensis Ivachnenko, 1997; Tokosauridae Tverdochlebova et Ivachnenko, 1984: Macroleter poezicus Tverdochlebova et Ivachnenko, 1984; Varanopidae Romer et Price, 1940: Mesenosaurus romeri Efremov, 1938; Pyozia mesensis Anderson et Reisz, 2004; Niaftasuchidae Ivachnenko, 1990: Niaftasuchus zekkeli Ivachnenko, 1990. Material examined: Niaftasuchus zekkeli Ivachnenko, 1990. Specimen PIN, no. 3717/36, partial skull (Fig. 61b). Ivakhnenko, 1990, text-fig. 3 (holotype); 2001, textfig. 35a; 2003c, p. 369, text-fig. 1; Ivakhnenko et al., 1997, pl. 70, fig. 2. Specimen PIN, no. 3717/39, cheek tooth (Fig. 41c). Ivakhnenko, 1990, text-fig. 1 (Rhopalodon sp.); Ivakhnenko et al., 1997, p. 78. Specimen PIN, no. 3717/40, cheek tooth. Ivakhnenko, 1990, p. 85 (Rhopalodon sp.); Ivakhnenko et al., 1997, p. 78. Collection PIN, no. 4276 (Golyusherma locality) Russian Federation, Udmurtia, Alnashskii District; Biarmian Series (Middle Permian), Upper Kazanian Substage; Ocher Faunal Assemblage, Golyusherma Faunal Subassemblage. Taphonomic association: Archegosauridae Meyer, 1858: Platyoposaurus sp.; Melosauridae Fritsch, 1885: Melosaurus compilatus Golubev, 1995; Leptorophidae gen. indet.; Archaeosyodontidae Ivachnenko, fam. nov.: Microsyodon orlovi Ivachnenko, 1995. Material examined: Microsyodon orlovi Ivachnenko, 1995. Specimen PIN, no. 4276/13, maxillary (Figs. 51a, 51b). Ivakhnenko, 1995a, text-figs. 4a and 4b (holotype); 2003c, p. 364, text-fig. 8; Ivakhnenko et al., 1997, pl. 71, fig. 1. Collection PIN, no. 4309 (Sokol locality) Russian Federation, Udmurtia, Zav’yalovskii District; Biarmian Series (Middle Permian), Upper KazaPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

875

nian Substage; Ocher Faunal Assemblage, Ocher Faunal Subassemblage. Taphonomic association: Eotitanosuchidae Tchudinov, 1960: Biarmosuchus tchudinovi Ivachnenko, 1999. Material examined: Biarmosuchus tchudinovi Ivachnenko, 1999. Specimen PIN, no. 4309/1, left maxillary (Figs. 50a, 50b). Ivakhnenko et al., 1997, p. 75 (Biarmosuchus sp.); Ivakhnenko, 1999, text-figs. 4b and 4c (holotype); 2003c, p. 368, text-fig. 7. Specimen PIN, no. 4309/2, right maxillary, possibly belonging to the same individual as PIN, no. 4309/1. Ivakhnenko et al., 1997, p. 75 (Biarmosuchus sp.); Ivakhnenko, 1999, p. 85; 2003c, p. 368. Collection PIN, no. 4312 (Sidorovy Gory locality) Russian Federation, Udmurtia, Votkinskii District; Biarmian Series (Middle Permian), Upper Kazanian Substage; Ocher Faunal Assemblage, Golyusherma Faunal Subassemblage. Taphonomic association: Melosauridae Fritsch, 1885: Melosaurus sp.; Phthinosuchidae Efremov, 1954: Kamagorgon ulanovi Tatarinov, 1999. Material examined: Kamagorgon ulanovi Tatarinov, 1999. Specimen PIN, no. 4312/1, partial skeleton (Figs. 44a, 66b). Ivakhnenko et al., 1997, p. 75 (Phthinosuchidae (?) gen. indet.); Tatarinov, 1999b, text-fig. 1 (holotype); Ivakhnenko, 2001, text-fig. 30a; 2003c, p. 368, text-fig. 26. Collection PIN, no. 4337 (Kichkass locality) Russian Federation, Orenburg Region, Perevolotskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Isheevo Faunal Assemblage, Malaya Kinel Faunal Subassemblage. Taphonomic association: Enosuchidae Konzhukova, 1955: Enosuchus cf. breviceps Konzhukova, 1955; Microuraniidae Ivachnenko, 1995: Microurania minima Ivachnenko, 1995; Deuterosauridae Seeley, 1894: Deuterosaurus sp.; Ulemosauridae Ivachnenko, 1994: Ulemosaurus cf. gigas (Efremov, 1954); Ulemicidae Ivachnenko, 1996: Ulemica sp. Material examined: Microurania minima Ivachnenko, 1995. Specimen PIN, no. 4337/1, partial skull (Figs. 6a, 6b, 52a–52d, 61d); Ivakhnenko, 1995b, text-fig. 2 (holotype); 2001, text-fig. 28; 2003c, p. 365, text-fig. 2; Ivakhnenko et al., 1997, pl. 70, fig. 1.

876

IVAKHNENKO

Collection PIN, no. 4416 (Ust’-Koin locality) Russian Federation, Komi Republic, Knyazhpogostskii District; Biarmian Series (Middle Permian), Upper Kazanian Substage; Ocher Faunal Assemblage, Golyusherma Faunal Subassemblage. Taphonomic association: Melosauridae Fritsch, 1885: Koinia silantjevi Gubin, 1993; Dissorophidae Boulenger, 1902: Alegeinosaurus sp.; Karpinskiosauridae gen. indet.; Captorhinidae Case, 1911: Riabininus cf. uralensis (Riabinin, 1915); Bolosauridae Cope, 1878: Timanosaurus ivachnenkoi Gubin, 1993; Rhopalodontidae Seeley, 1894: Parabradysaurus silantjevi Ivachnenko, 1995; Phthinosuchidae Efremov, 1954: Kamagorgon sp.; Alrausuchidae gen. indet. Material examined: Parabradysaurus silantjevi Ivachnenko, 1995. Specimen PIN, no. 4416/4, cheek tooth (Fig. 43c). Ivakhnenko, 1995b, p. 115; 2003c, p. 368; Ivakhnenko et al., 1997, pl. 67, fig. 5a. Specimen PIN, no. 4416/9, incisor (fig. 43b). Ivakhnenko, 1995b, p. 115; 2003c, p. 368; Ivakhnenko et al., 1997, p. 77. Specimen PIN, no. 4416/33, partial dentary. Ivakhnenko, 1995b, text-fig. 1b (holotype); 2001, text-fig. 28a; 2003c, p. 368, text-fig. 35b; Ivakhnenko et al., 1997, pl. 67, fig. 5. Specimen PIN, no. 4416/36, partial maxillary (Fig. 43a). Ivakhnenko, 1995b, p. 115; 2003c, p. 368, text-fig. 35b; Ivakhnenko et al., 1997, p. 77. Kamagorgon sp. Specimen PIN, no. 4416/41, upper canine. Ivakhnenko et al., 1997, p. 77; Ivakhnenko, 2003c, p. 369. Alrausuchidae gen. indet. Specimen PIN, no. 4416/46, right parietal. Collection PIN, no. 4538 (Vozdvizhenka locality) Russian Federation, Orenburg Region, Asekeevskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Isheevo Faunal Assemblage. Taphonomic association: Melosauridae gen. indet.; Karpinskiosauridae Sushkin, 1925: Karpinskiosaurus sp.; Procolophonia fam. indet.; Nycteroleteridae gen. indet.; Microuraniidae Ivachnenko, 1995: Microurania mikia Ivachnenko, 2003. Material examined:

Microurania mikia Ivachnenko, 2003. Specimen PIN, no. 4538/7, partial dentary (Fig. 52e). Ivakhnenko, 2003c, p. 370, text-fig. 3 (holotype, incorrectly designated as no. 4338/7). Specimen PIN, no. 4538/29, caniniform tooth (Fig. 52f). Collection PIN, no. 4539 (Ibryaevo locality) Russian Federation, Orenburg Region, Krasnogvardeiskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Malaya Kinel Faunal Assemblage. Taphonomic association: Melosauridae Fritsch, 1885: Tryphosuchus sp.; Rhopalodontidae Seeley, 1894: Phthinosaurus sp. Material examined: Phthinosaurus sp. Specimen PIN, no. 4539/1, partial right maxillary (Figs. 42d, 42e). Ivakhnenko et al., 1997, p. 62. Three cheek teeth are preserved; their crowns are very unusual, most similar to those of Parabradysaurus, but have a weak serration of borders. At the base of one tooth, there is a partial crown of a replacement tooth, which is very similar to the crown of the replacement tooth of the type specimen of Phthinosaurus borissiaki; this was the reason to assign it to this genus. Collection PIN, no. 4541 (Ust’-Vashka locality) Russian Federation, Arkhangelsk Region, Leshukonskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Mezen Faunal Assemblage. Taphonomic association: Nyctiphruretidae Efremov, 1938: Nyctiphruretus acudens Efremov, 1938; Nycteroleteridae Romer, 1956: Bashkyroleter mesensis Ivachnenko, 1997; Tokosauridae Tverdochlebova et Ivachnenko, 1984: Macroleter poezicus Tverdochlebova et Ivachnenko, 1984; Varanopidae Romer et Price, 1940: Mesenosaurus romeri Efremov, 1938; Nikkasauridae Ivachnenko, 2000: Nikkasaurus tatarinovi Ivachnenko, 2000; Reiszia tippula Ivachnenko, 2000; Alrausuchidae fam. nov.: Alrausuchus tagax (Ivachnenko, 1990); Niaftasuchidae Ivachnenko, 1990: Niaftasuchus zekkeli Ivachnenko, 1990. Material examined: Reiszia tippula Ivachnenko, 2000. Specimen PIN, no. 4541/2, partial left dentary (Fig. 40e). Ivakhnenko, 2000b, text-fig. 6 (holotype). PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

Niaftasuchus zekkeli Ivachnenko, 1990. Specimen PIN, no. 4541/12, partial left dentary. Ivakhnenko et al., 1997, p. 77. Collection PIN, no. 4543 (Nisogora locality) Russian Federation, Arkhangelsk Region, Leshukonskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Mezen Faunal Assemblage. Taphonomic association: Nyctiphruretidae Efremov, 1938: Nyctiphruretus acudens Efremov, 1938; Tokosauridae Tverdochlebova et Ivachnenko, 1984: Macroleter poezicus Tverdochlebova et Ivachnenko, 1984; Lanthaniscidae Ivachnenko, 2007: Lanthaniscus efremovi Ivachnenko, 1980; Caseidae Williston, 1912: Ennatosaurus tecton Efremov, 1956; Varanopidae Romer et Price, 1940: Mesenosaurus romeri Efremov, 1938; Niaftasuchidae Ivachnenko, 1990: Niaftasuchus zekkeli Ivachnenko, 1990. Material examined: Niaftasuchus zekkeli Ivachnenko, 1990. Specimen PIN, no. 4543/20, partial right dentary (Figs. 41a, 61b). Ivakhnenko et al., 1997, p. 70. Collection PIN, no. 4548 (Ust’e Strel’ny locality) Russian Federation, Vologda Region, Velikoustyugskii District; Tatarian Series (Upper Permian), Upper Severodvinian Substage, Severodvinian Horizon; Sokolki Faunal Assemblage, Ilinskoe Faunal Subassemblage. Taphonomic association: Dvinosauridae Amalitzky, 1921: Dvinosaurus sp.; Kotlassiidae Romer, 1934: Microphon exiguus Ivachnenko, 1983; Chroniosuchidae Vjuschkov, 1957: Chroniosaurus dongusensis Tverdochlebova, 1972; Gorgonopidae Lydekker, 1890: Suchogorgon golubevi Tatarinov, 2000; Galeopidae Broom, 1912: Suminia sp.; Ictidorhinidae Broom, 1932: Ustia atra Ivachnenko, 2003. Material examined: Suchogorgon golubevi Tatarinov, 2000. Specimen PIN, no. 4548/1, skull. Ivakhnenko et al., 1997, p. 78 (Sauroctonus aff. progressus); Tatarinov, 2000, pp. 70–78, text-figs. 1–3 (holotype); Ivakhnenko, 2003c, pp. 369, 399; Ivakhnenko, 2005b, p. 401, textfig. 1. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

877

Specimen PIN, no. 4548/10, skull. Ivakhnenko et al., 1997, p. 78 (Sauroctonus aff. progressus); Tatarinov, 2000, pp. 70–78, text-fig. 4; Ivakhnenko, 2003c, p. 369; Ivakhnenko, 2005b, p. 402, text-figs. 2 and 20. Specimen PIN, no. 4548/58, upper canine (Fig. 45b). Ivakhnenko, 2005b, p. 402, text-fig. 33c. Specimen PIN, no. 4548/138, partial skull. Ivakhnenko, 2003c, p. 369; Ivakhnenko, 2005b, p. 402, textfig. 3. Specimen PIN, no. 4548/158, mandible (Fig. 45a). Ivakhnenko, 2003b, text-fig. 5 (incorrectly designated as SGU, no. 104/1767). Specimen PIN, no. 4548/161, cheek tooth (Fig. 45c). Ivakhnenko, 2005b, p. 403. Specimen PIN, no. 4548/163, left frontal (Fig. 3a). Ivakhnenko, 2005b, p. 403. Specimen SGU, no. 104B/1767, partial skull. Ivakhnenko, 2003a, p. 50, text-figs. 2b–2g, 3, and 4; 2003c, p. 369; 2005b, p. 403. Ustia atra Ivachnenko, 2003. Specimen PIN, no. 4548/155, right dentary (Figs. 45a, 61c). Ivakhnenko, 2003c, pp. 369, 394, textfig. 21b (holotype). Suminia sp. Specimen SGU, no. 104B/1351, right premaxilla (Fig. 56a). Ivakhnenko et al., 1997, p. 78 (Suminia cf. getmanovi). Collection PIN, no. 4549 (Klimovo-1 locality) Russian Federation, Vologda Region, Velikoustyugskii District; Tatarian Series (Upper Permian), Lower Vyatkian Substage, Vyatkian Horizon; Sokolki Faunal Assemblage. Taphonomic association: Dvinosauridae Amalitzky, 1921: Dvinosaurus sp.; Kotlassiidae gen. indet.; Rubidgeidae Broom, 1938: Leogorgon klimovensis Ivachnenko, 2003; Dicynodontidae gen. indet.; Procynosuchidae Broom, 1937: Uralocynodon sp. Material examined: Leogorgon klimovensis Ivachnenko, 2003. Specimen PIN, no. 4549/13, partial periotic; Ivakhnenko, 2003c, p. 365, text-fig. 28 (holotype). Dicynodontidae gen. indet. Specimen PIN, no. 4549/3, right maxillary (Fig. 9b). Specimen PIN, no. 4549/23, dentaries. Collection PIN, no. 4639 (Belokur’e locality) Russian Federation, Arkhangelsk Region, Mezenskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Mezen Faunal Assemblage. Taphonomic association:

878

IVAKHNENKO

Nyctiphruretidae Efremov, 1938: Nyctiphruretus acudens Efremov, 1938; Nycteroleteridae Romer, 1956: Bashkyroleter mesensis Ivachnenko, 1997; Tokosauridae Tverdochlebova et Ivachnenko, 1984: Macroleter poezicus Tverdochlebova et Ivachnenko, 1984; Varanopidae Romer et Price, 1940: Mesenosaurus romeri Efremov, 1938; Nikkasauridae Ivachnenko, 2000: Nikkasaurus tatarinovi Ivachnenko, 2000. Material examined: Nikkasaurus tatarinovi Ivachnenko, 2000. Specimen PIN, no. 4639/1, partial skeleton. Ivakhnenko et al., 1997, p. 56 (Therapsida, ord. indet.); Ivakhnenko, 2000b, text-fig. 2. Collection PIN, no. 4644 (Voskresenskoe-2B locality) Russian Federation, Nizhni Novgorod Region, Voskresenskii District; Tatarian Series (Upper Permian), Upper Vyatkian Substage, Vyatkian Horizon; Sokolki Faunal Assemblage, Sokolki Faunal Subassemblage. Taphonomic association: Dvinosauridae Amalitzky, 1921: Dvinosaurus sp.; Chroniosuchidae Vjuschkov, 1957: Chroniosuchus sp.; Dicynodontidae Owen, 1859: Delectosaurus arefjevi Kurkin, 2001. Material examined: Delectosaurus arefjevi Kurkin, 2001. Specimen PIN, no. 4644/1, skull (Figs. 5a, 5b, 9a, 14b). Kurkin, 2001, text-figs. 1 and 2 (holotype); Ivakhnenko, 2003c, text-fig. 50. Collection PIN, no. 4659 (Kosmogorodsko’e locality) Russian Federation, Arkhangelsk Region, Mezenskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Mezen Faunal Assemblage. Taphonomic association: Nyctiphruretidae Efremov, 1938: Nyctiphruretus acudens Efremov, 1938; Nycteroleteridae Romer, 1956: Bashkyroleter mesensis Ivachnenko, 1997; Tokosauridae Tverdochlebova et Ivachnenko, 1984: Macroleter poezicus Tverdochlebova et Ivachnenko, 1984; Varanopidae Romer et Price, 1940: Mesenosaurus romeri Efremov, 1938; Alrausuchidae fam. nov.: Alrausuchus tagax (Ivachnenko, 1990). Material examined: Alrausuchus tagax Ivachnenko, 1990.

Specimen PIN, no. 4659/8, partial skull (Figs. 22b, 49a–49d, 64a). Collection PIN, no. no. 4660 (Dorogaya Gora locality) Russian Federation, Arkhangelsk Region, Mezenskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Mezen Faunal Assemblage. Taphonomic association: Nyctiphruretidae Efremov, 1938: Nyctiphruretus acudens Efremov, 1938; Nycteroleteridae Romer, 1956: Bashkyroleter mesensis Ivachnenko, 1997; Tokosauridae Tverdochlebova et Ivachnenko, 1984: Macroleter poezicus Tverdochlebova et Ivachnenko, 1984; Varanopidae Romer et Price, 1940: Mesenosaurus romeri Efremov, 1938; Niaftasuchidae Ivachnenko, 1990: Niaftasuchus zekkeli Ivachnenko, 1990. Material examined: Niaftasuchus zekkeli Ivachnenko, 1990. Specimen PIN, no. 4660/31, cheek tooth (Fig. 41b). Collection PIN, no. 4678 (Port Kotelnich locality) Russian Federation, Kirov Region, Kotelnichskii District; Tatarian Series (Upper Permian), Upper Severodvinian Substage, Severodvinian Horizon; Sokolki Faunal Assemblage, Kotelnich Faunal Subassemblage. Taphonomic association: Bradysauridae Huene, 1948: Deltavjatia vjatkensis (Hartmann-Weinberg, 1937); Phthinosuchidae Efremov, 1954: Viatkogorgon ivakhnenkoi Tatarinov, 1999; Pristerodontidae Toerien, 1953: Australobarbarus kotelnitshi Kurkin, 2000; Australobarbarus platycephalus Kurkin, 2000. Material examined: Australobarbarus kotelnitshi Kurkin, 2000. Specimen PIN, no. 4678/2, skull (Figs. 39d, 68a). Kurkin, 2000, text-fig. 1 (holotype); Ivakhnenko, 2003c, text-fig. 44. Australobarbarus platycephalus Kurkin, 2000. Specimen PIN, no. 4678/3, skull. Kurkin, 2000, text-fig. 2 (holotype); Ivakhnenko, 2003c, text-fig. 45. Australobarbarus sp. Specimen PIN, no. 4678/8, left dentary (Figs. 45a, 45b). Viatkogorgon ivakhnenkoi Tatarinov, 1999. Specimen PIN, no. 4678/5, skull. Ivakhnenko et al., 1997, p. 72 (Gorgonopidae gen. indet.); Ivakhnenko, 2003c, p. 367, text-fig. 27b. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

Collection SGU, no. 104B (Borisov locality) Russian Federation, Orenburg Region, Buguruslanskii District; Biarmian Series (Middle Permian), Urzhumian Stage, Urzhumian Horizon; Ocher Faunal Assemblage, Ocher Faunal Subassemblage. Taphonomic association: Rhopalodontidae Seeley, 1894: Rhopalodon sp. Material examined: Rhopalodon sp. Specimen PIN, no. 2050, right dentary (Fig. 42b). Ivakhnenko et al., 1997, p. 57 (Rhopalodon (?) sp.); Ivakhnenko, 2003c, p. 363. Collection SGU, no. 104B (Dubovka-1 locality) Russian Federation, Orenburg Region, Novosergeevskii District; Tatarian Series (Upper Permian), Upper Severodvinian Substage, Severodvinian Horizon; Sokolki Faunal Assemblage, Ilinskoe Faunal Subassemblage. Taphonomic association: Ictidorhinidae Broom, 1932: Biarmosuchoides romanovi Tverdochlebova et Ivachnenko, 1994. Material examined: Biarmosuchoides romanovi Tverdochlebova et Ivachnenko, 1994. Specimen PIN, no. 2051, dentary. Tverdochlebova and Ivachnenko, 1994, text-figs. 2c and 2d (holotype); Ivakhnenko et al., 1997, pl. 67, fig. 1; Ivakhnenko, 2003c, p. 363, text-fig. 21a. Collection TsNIGR Museum, no. 1727 (Mezhevaya locality) Russian Federation, Udmurtia, Sarapul’skii District; Biarmian Series (Middle Permian), Lower Kazanian Substage; Ocher Faunal Assemblage, Golyusherma Faunal Subassemblage. Taphonomic association: Rhopalodontidae Seeley, 1894: Parabradysaurus udmurticus Efremov, 1954. Material examined: Parabradysaurus udmurticus Efremov, 1954. Specimen PIN, no. 2/1727, partial dentary. Efremov, 1954, text-fig. 69; Tchudinov, 1983, text-fig. 28; Ivakhnenko et al., 1997, pl. 67, fig. 4; Ivakhnenko, 2003c, p. 366, text-fig. 35a. CHAPTER 3. COMPARATIVE CRANIAL MORPHOLOGY As seen in the list of the material investigated, the majority of Dinomorpha groups are represented in East European localities; some specimens are extraordinarily well preserved. This provides the opportunity to examine many interesting structural features. Certainly, PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

879

it is impossible to describe in one work all structural variants of skull bones in this diverse group, even taking into account only the East European collections. Therefore, I intend to restrict myself to the most contrasting examples, the structures which have certain functional interpretations, and structural features that have not previously been remarked in the literature. The skull of Suchogorgon, which is particularly thoroughly examined (Ivakhnenko, 2005b), is taken as reference material. It is expedient to describe cranial structures following the segments with certain functions rather than according to the genetic principle, which is usually accepted (and which was used in the study of the skull of a particular taxon, see Ivakhnenko, 2005b). This will considerably facilitate further interpretation of these structures. The following scheme of comparative description of cranial structures is accepted: (1) Skull roof (ossa calvaria). The major function of this region is to form the cover of the nasal capsule and braincase; in this section, available data on sclerotic bones are considered. (2) Cheek segment (ossa buccalia). The major function is connected with the upper jaw. (3) Temporal segment (ossa temporalia). The major function is connected with the jaw muscles. (4) Palatal segment (ossa palatalia). The major function is connected with the separation of the alimentary and respiratory systems. (5) Lower jaw (ossa mandibularia). The major function is connected with the articulation of the jaw apparatus. (6) Ethmoid and cerebral region (endocranium). Endochondral cartilage bones are considered. (7) Auditory structures. (8) Dentition (dentes). The system is connected with food treatment. (9) Ontogenetic changes in skull. Theoretically, this topic is of great importance for the reconstruction of phylogeny. However, available data allow the differentiation of taxonomic and age differences in only a few cases. 1. Skull Roof (ossa calvaria) The dorsal cover of the braincase and nasal capsule display a standard general design. It is usually composed of the nasals, frontals, prefrontals, postfrontals, lachrymals, parietals, tabulars, and postparietals. In some cases, the preparietal is added; the postfrontals (in some Dicynodontida) are sometimes reduced or fused with adjacent bones. The supratemporals, which are always present in Sphenacomorpha, are characteristic of the majority of Dinomorpha. The ventral (internal) surface of the bones overlying the endocranium is smooth, small foramina for blood vessels are located at the ossification centers. Small

880

IVAKHNENKO (a)

L Smx

Prf

Pmx

(b) L cl sh cif

adl Smx

Fig. 1. Nasal cavities, sagittal section, intracranial view: (a) Suminia getmanovi Ivachnenko, 1994, specimen PIN, no. 2212/62; and (b) Ulemica invisa (Efremov, 1938), specimen PIN, no. 157/1116. Designations: (adl) apertura anterior of the lachrymal duct, (cif) crista infralacrimalis of the lachrymal, (cl) crista lacrimalis of the lachrymal, (L) lachrymal, (pim) processus intermedialis of the septomaxillary, (Pmx) premaxilla, (Prf) praefrontal, (sh) sinus maxillaris (sinus Highmori), and (Smx) septomaxillary. Scale bar, 1 cm.

areas are located on each side anterior to the parietal foramen. This area (facies epipterygoidei) is articulated with a thin epiptrygoid (forming the sutura sphenoparietalis). Anterior to the areas, the surface of the nasals and frontals has cristae cranii, which extend at the midwidth of bones and mark the area of contact of the roof of the ethmoid structures and the olfactory region. In all taxa examined, this region is similar in design (Ivakhnenko, 2005b, text-fig. 21); the minor differences are restricted to certain variation in the extent to which the cristae are developed. In all taxa investigated, the internal surface of the prefrontal is concave and forms the posterodorsal corner (recessus praefrontalis) of the ethmoturbinal region between the ethmoid tube and the skull wall (Ivakhnenko, 2005b, text-fig. 26). The internal (medial) surface of the lachrymal is more complex. The anterodorsal margin of the bone passes anteriorly as a thin plate, covering from within the upper margin of the facial lamina of the maxillary and the prefrontal–maxillary suture. In canineless ano-

modonts (Suminia, Ulemica: Figs. 1a, 1b), this plate passes over the internal surface of the maxillary, closely approaching the suture with the septomaxillary. The overlying plate contains the lachrymal duct, which forms a rounded crest on the intracranial surface (crista lacrimalis: Fig. 1a, cl). In taxa that have a well-developed canine sheath (Biarmosuchus, Titanophoneus, Syodon, Inostrancevia, Estemmenosuchus, and Suchogorgon: Ivakhnenko, 2005b, text-fig. 26), the duct raises anteriorly and, close to the uppermost point, the facial lamina of the maxillary has a foramen, the apertura anterior ductus lacrimalis (Fig. 1a, adl). In canineless taxa, the duct is almost horizontal or even slightly descending. A thin crest (crista infralacrimalis: Fig. 1a, cif) extends anteriorly from the posterior margin of the crista lacrimalis and passes into the crista infralacrimalis of the maxillary, separating the sinus maxillaris Highmori from the lower olfactory cavity. Bones of the skull roof are mostly connected by serrated sutures. The serration is usually most pronounced PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA) Prf

(a)

881

(b)

fl

ffp

Po fep

Fig. 2. Structures of the skull roof bones of Estemmenosuchus uralensis Tchudinov, 1960: (a) right parietal, sagittal view, specimen PIN, no. 1758/325; and (b) frontoparietal tubercle of a young individual, specimen PIN, no. 1758/80. Designations: (fep) facies epipterygoidei of the parietal, (ffp) facies frontalis of the parietal, (fl) facies postparietalis of the parietal, (Po) postorbital, and (Prf) prefrontal. Scale bar, 1 cm.

in the coronal suture between the frontal and parietal. Judging from the bone topography, the postfrontal was fused with the frontal in Idelesaurus (Ivakhnenko, 2003c, text-fig. 46b) and, according to the data of Cluver and King (1983), in Endothiodon, Rhachiocephalus, Aulacephalodon, Pelanomodon, and Kingoria. The sagittal suture is the flattest and has weak, obliquely longitudinal crests. This probably provided additional reinforcement of the suture. For example, in large individuals of Biarmosuchus, the region of the coronal suture shows superficial obliteration. In the case of well-pronounced pachyostosis (Ulemosaurus, Deuterosaurus, Titanophoneus, Proburnetia, and Estemmenosuchus), the suture is almost completely obliterated. However, this is not universally true; for example, in Estemmenosuchus (Fig. 2a), very thick bones are combined with a weak suture in this region. In the case of well-developed pachyostosis, for example, in Proburnetia, not only the sagittal suture, but also the frontonasal suture and even the coronal suture are obliterated, although traces of sutures are usually retained on the ventral surface of the bones. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

In Inostrancevia, the sagittal suture is reinforced by complex bends, which extend far laterally (Ivakhnenko, 2000a, text-fig. 6). A special pattern of strengthening the suture is the formation in the middle part of the coronal suture of particular ossifications, such as the preparietal (interparietal). This superficial ossification is probably formed of wormian bones of the sagittal suture. Thus, one specimen of Prorubidgea (Sigogneau, 1970a, pl. 68a) has a small irregular bone in this region. Only in Gorgonopidae and Dicynodontida, the bone is always present. The independent development of this bone in the two groups is evident from the fact that, in the first, it does not reach the parietal foramen, while, in the second, it forms the anterior border of this foramen. In Dinocephalia, the parietal region of the skull roof is more or less narrowed anterior to the parietal foramen; this is connected with the development of depressions for jaw muscles, which are primarily limited by the postorbital crest of the postorbital. This narrowing is particularly well pronounced in Syodon and Deuterosaurus, in which the area of parietals not covered by muscles is limited to the margins of the tubercle sur-

882

IVAKHNENKO

rounding the parietal foramen. The parietals of Dicynodontia are even more narrowed; they are longitudinally elongated, sometimes forming a parietal crest (e.g., in Delectosaurus and many others). The primary connection of the parietal region to the adjacent functional segments (cheek and temporal skull regions) is relatively weak. The cheek region or, more precisely, the facial lamina of the maxillary of primitive taxa (Nikkasaurus, Niaftasuchus, and Microurania) overlies flat margins of the nasal, prefrontal, and lachrymal. A simple connection is retained in many predatory taxa, such as Archaeosyodon, Syodon, Titanophoneus, and Dinosaurus. The weakest connection in this region is characteristic of saber-toothed predators (Biarmosuchus, Alrausuchus, Inostrancevia, Pravoslavlevia, Sauroctonus, and Suchogorgon); this suggests that certain limited mobility of the maxillary had functional significance (Ivakhnenko, 2005b, p. 439). The articular surfaces of bones in this region lack even weak crests. The flat surfaces display only weak, branching grooves of blood vessels. The bones were probably connected by a layer of connective tissue. In relatively large phytophagous taxa (Ulemosaurus, Estemmenosuchus, etc.) and all Anomodontia investigated, the bones are connected by more rigid sutures, which almost reach a serrated state. The connection with the temporal region is also weak. In all taxa with the apopareial skull pattern, a weak zone of the former spiracular fissure is retained between the bones of the roof and cheek regions. In Eotherapsida, this zone passes along the caudal process of the postorbital, reaching the squamosal. Below this process, there is the synapsid fenestra; above the process, there is a zone of the parapsid fissure that is not open. In primitive designs (in this case, above the level of Sphenacomorpha), this region is bordered from the side of the roof by the parietal, supratemporal (if it is retained), and the tabular. The lateral margin of the parietal is thin, widely overlies the caudal process of the postorbital without a suture (facies postorbitalis). The supratemporal has been recorded in Suchogorgon (Ivakhnenko, 2005b, text-figs. 7, 8); this bone is probably retained in all Gorgonopidae and maybe present in Biarmosuchus; at least, the left side of specimen PIN, no. 1758/85 retains a fragment of a thin plate (probably apophysis paroccipitalis ossis supratemporalis), which distinctly separates the squamosal from the paroccipital process. In fact, this bone is probably homologous to the squamosal (or squama temporalis ossis temporalis) of mammals (Ivakhnenko, 2002a, 2003c, 2005b). The ala ascendens of this bone overlies the parietal (Ivakhnenko, 2005b, text-fig. 21), forming the squama temporalis. The bone is probably retained in other taxa, but its thin plate is probably included in the multilayer “package” of the posteromedial part of the wall of the cavum temporalis, which is characteristic of all groups of Dinomorpha and is formed of many bones (squamosal,

postparietal, parietal, supratemporal, and tabular); this probably results in reduction of this bone. At least, it is absent from higher Dicynodontia. In Suchogorgon (Ivakhnenko, 2005b, text-fig. 8), the lateral margin of the bone descends vertically in the shape of a parotic lamina and separates the upper part of the paroccipital process from the squamosal. The contact surfaces of the parotic lamina and squamosal are flat, without a pronounced suture. The tabular comes in contact with the squamosal only in the region of the mastoid crest; however, the contact surfaces are also flat. In dicynodonts (Idelesaurus, Dicynodon, and Delectosaurus), the mastoid crest is absent, and the tabular is very short, terminates short of reaching the posttemporal fenestra. The bone is connected to the postparietal (here, the postparietals are fused in a single bone) by a serrated suture and its long projections even overlie the surface of the endochondral supraoccipital. In all taxa, at the external margin, the tabular forms an only slightly complicated connection with the squamosal; the two bones are slightly thickened along the suture, although the internal surface of the suture remains smooth. On the dorsal side of the skull roof, the surface sculpture of bones is of particular interest. Unfortunately, details of the sculpture are usually indistinct because of unsatisfactory preservation or inappropriate preparation. Therefore, only poor data are presently available. In small primitive taxa (Nikkasaurus, Niaftasuchus, Otsheria, Ulemica, Suminia) and taxa with thin bones (Alrausuchus, Biarmosuchus, Syodon, Inostrancevia), the surface is almost smooth, with small foramina and grooves extending from them at the centers of ossification. Taxa with thick bones (Archaeosyodon, Pravoslavlevia, Suchogorgon, Sauroctonus) show a background sculpture, which is particularly distinctly pronounced in strongly thickened, but nonpachyostotic bones (Kamagorgon). The bone surfaces are covered with small irregularly arranged tubercles frequently fused in rough varicoid ridges. Small foramina open between them and sometimes give rise to narrow grooves. As the bone surface is very well preserved, Suchogorgon displays an unusual sculpture on the frontal (Ivakhnenko, 2005b, p. 415, text-fig. 20). In the middle part of the bone, somewhat closer to the orbital margin, there is a field of deep circular or ovate fossae from 2 to 3 mm in diameter; sometimes, two or three fossae are fused to form a deep ovate depression. The bottom of the fossae is uniformly rough rather than smooth. A large lacuna inside the bone (Fig. 3a) opens on the surface in large foramina, which usually give rise to branching grooves on the bone surface. The foramina sometimes coincide with fossae, but usually lie at their borders, although may open between fossae. These blood vessels are probably related to the fossae, since in the case that vessels do not open in a fossa, it is always connected by a groove with the nearest fossa with a foramen for vessels. In contrast to the rough bottom of the fossae, the bottom of grooves for vessels is always PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

883

(a)

(b) F

Pof

P Fig. 3. Structures of the frontal: (a) Suchogorgon golubevi Tatarinov, 2000, longitudinal section through the posterior part of the bone, specimen PIN, no. 4548/163; and (b) Alrausuchus tagax (Ivachnenko, 1990), part of the roof, dorsal view, holotype PIN, no. 3706/10. Designations: (F) frontal, (P) parietal, and (Pof) postfrontal. Scale bar, 1 cm.

smooth. Similar structures are observed in all Gorgonopidae investigated and probably characteristic of many other taxa. At least, Alrausuchus (Fig. 3b) and Biarmosuchus, with thin frontals and very weak background sculpture, have in this place a distinct elongated depression, which contains foramina for large vessels also coming from the blood lacuna inside the bone. Well developed branching grooves extend from these foramina. Similar structures are observed in a young individual of Titanophoneus (PIN, no. 157/180, right frontal). A deep and wide depression with a rough and coarse bottom is observed here; it contains a series of large foramina giving rise to large branching grooves for blood vessels. In large individuals of Titanophoneus, the grooves pass onto a large frontoparietal pachyostotic expansion. In Proburnetia, vessels coming from the same region twined around hornlike outgrowths on the parietal and above the orbits (Fig. 4a). Pachyostotic expansions in the frontoparietal region are characteristic of large dinocephals, such as Ulemosaurus and Titanophoneus. As pachyostotic expansions were formed, canals for blood vessels passed inside the bone and, then, branched and anastomosed on the surface. In Proburnetia, pachyostotic thickenings surround the parietal foramen and extend anteriorly in the PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

shape of a narrow crest, passing onto the nasal (Ivakhnenko, 2003c, text-figs. 32a, 32b); in Deuterosaurus, they are only developed above the orbits and on the nasals (Ivakhnenko, 2003c, text-figs. 18a, 18b). In these cases, vessels passed onto pachyostotic tubercles and branched. In young Estemmenosuchus, the surface of pachyostotic expansions is also covered with the texture of vessels (Fig. 2b), while, in presumable adult males (Ivakhnenko, 2000a), tubercles are transformed into long, sometimes branching hornlike outgrowths with an almost smooth surface (Fig. 4b). These structures on the surface of the skull roof are probably connected with extensive vascularization and presumably associated with primitive thermoregulation (Ivakhnenko, 2005b, p. 440). The sculpture on the skull roof is not characteristic of Anomodontia. The bones of this region are usually thin and almost smooth, with small foramina for blood vessels and grooves deviating from them (Fig. 5a). The group as a whole lacks pachyostotic expansions in the frontoparietal region. However, some taxa, such as Dicynodon trautscholdi, Delectosaurus, and Interpresosaurus, have well-pronounced circular tubercles in the anterior or middle part of the nasals; according to Cluver and King (1983), the same is true of Oudeno-

884

IVAKHNENKO

(a) (b)

Fig. 4. Pachyostotic outgrowths of the left side of the frontoparietal region of the skull: (a) Proburnetia vjatkensis Tatarinov, 1968, holotype PIN, no. 2416/1; and (b) Estemmenosuchus mirabilis Tchudinov, 1968, holotype PIN, no. 1758/6. Scale bar, 2 cm.

(a)

(b)

Fig. 5. Dorsal surfaces of cranial bones of Delectosaurus arefjevi Kurkin, 2001, holotype PIN, no. 4644/1, right side: (a) frontal and preparietal and (b) nasal. Scale bar, 1 cm. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

885

(a)

N (b) Smx F

Prf

Pof Mx

Fig. 6. Membrane bones of the skull roof of Microurania minima Ivachnenko, 1995, holotype PIN, no. 4337/1, dorsal view: (a) frontonasal region; and (b) septomaxillary bone and anterior part of the maxillary bone of the right side. Designations: (F) frontal, (Mx) maxillary, (N) nasal, (Pof) postfrontal, (Prf) prefrontal, and (Smx) septomaxillary. Scale bar, 0.5 cm.

don, Aulacephalodon, and Pelanomodon. These formations are particularly characteristic of Triassic Kannemeyeriidae. Inside the bony tissue of these tubercles, there are many canals for blood vessels, which are much more abundant than is usual for the center of ossification of the nasals (Fig. 5b). These tubercles are very similar in structure to thickenings in the same region of extant Rhinocerotidae, in which similar tubercles form the base of the horn of fused guard hairs. Certainly, this question requires additional study. The sculpture on the nasals of small Microurania is unusual. The parietals and frontals have a usual weak background sculpture, while the nasals are covered with large, circular, densely spaced fossae, which pass onto the facial plates of the maxillaries (Figs. 6a, 6b). All Dinomorpha retain the parietal foramen. Its borders are usually elevated (except for the majority of Dicynodontida). The parietal foramen is often surrounded by a high tubercle with a rough, scalloped margin. In small primitive Nikkasaurus, the parietal foramen is located between the parietals and frontals. The parietal part is elevated, with a scalloped margin, while the frontal part is even, but also has a sharply scalloped margin. In Dicynodontida investigated, the border of the foramen is smooth. It is interesting that, in Dicynodontida, the border of the orbit is almost smooth or, sometimes, slightly rugose. In other taxa under study, rugosity on the orbital rim is much more developed; sometimes (in Microurania), it is almost scalloped, often sharply thickened and even pachyostotic (many PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

Dinocephalia). The anterior and upper borders of the circumorbital ring are formed by the lachrymal, prefrontal, postfrontals, and, to some extent, by the margin of the frontal. The extent to which the frontal contributes to the orbital rim widely varies even in closely related taxa; this variation is sometimes connected with relative sizes of orbits in young and old individuals. It is important to consider the sclerotic ring, a structure inside the orbit. It is poorly described in the literature, information is restricted to the presence of this structure in some taxa. The ring (annulus sclerotici) developed in the fibrous envelope of the eye and was probably related to the acuity of eyesight. Unfortunately, I am unaware of works connecting structural details of the ring with its functions. However, representatives of Dinomorpha differ essentially in the structure of the ring (see below). The ring has been recorded in the following East European taxa: Biarmosuchus, Alrausuchus, Titanophoneus, Nikkasaurus, Dinosaurus, Viatkogorgon, and Proburnetia. Among South African taxa, the presence of the ring was indicated in Ictidorhinus (Sigogneau, 1970a, text-fig. 191). The sclerotic ring has not been recorded in Anomodontia, except for Galechirus (Broom, 1907); however, it should be noted that, in the closely related East European Suminia, it is not found, although a large number of specimens have been examined. Sometimes, in the case of insufficient preparation of specimens, fragments of the orbitosphenoid or ala parasagittalis of the pterygoid are mistaken for bones of the sclerotic ring. In Biarmosuchus, Dinosaurus, and Proburnetia, the

886

IVAKHNENKO (a)

–

–

(b)

+

(c) + – –

+ (d) – +

+ +

–

–

Fig. 7. Sclerotic rings, right side, with the anterior edge on the right: (a) Nikkasaurus tatarinovi Ivachnenko, 2000, holotype PIN, no. 162/33; (b) Viatkogorgon ivakhnenkoi Tatarinov, 1999, holotype PIN, no. 2212/61; (c) Titanophoneus potens Efremov, 1938, lectotype PIN, no. 157/1; and (d) Alrausuchus tagax (Ivachnenko, 1990), holotype PIN, no. 3706/10. Signs (+) and (–) designate positive and negative locking elements. Scale bars: (a) 0.5 cm and (b–d) 1 cm.

ring is preserved as hardly discernible imprints and bone fragments; therefore, it is difficult to gain an understanding of its structure. In all taxa investigated, the sclerotic ring consists of four quadrants of intermediate plates separated by opposed positive and negative locking plates. The positive plates are located at the anterodorsal and posteroventral borders. The negative plates are at the posterodorsal and anteroventral borders. The intermediate plates are almost elongated rectangular. On one side (negative, i.e., facing the negative locking plate) of the internal border (base), there is a narrow triangular projection and, above it, there is a wider triangular depression. On the opposite side (positive), the internal border has a narrow depression and a wider triangular projection above it. Adjacent plates are articulated with slight

mobility by projections which overlap laterally the depressions. At the base of the positive plates, there are two narrow opposed projections and two wider depressions above them; the negative plates have two wide projections and two narrow depressions below them. The external margin of the plate is slightly convex and the base is slightly concave. At the same time, the external margin curves somewhat medially, while the base curves slightly laterally; hence, the plates curve externally (laterally). The upper plates are usually low, increasing in height downwards and, hence, the opening of the ring is displaced from the center. In Nikkasaurus (Fig. 7a, specimen PIN, no. 162/33), the ring is partially preserved in the left orbit, including several complete and about ten fragmentary plates. However, they overlap each other, and it is only possiPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

ble to examine the structure of the posterodorsal part of the ring. Here, there are an upper negative plate and two incomplete intermediate plates. The plates are considerably elongated and have small additional rounded triangular projections close to the middle of the positive margin. A locking plate has two projections of this kind. The plates are slightly convex, almost flat. Judging from the greatest mean width of elements and maximum possible diameter of the ring, it consisted of at most 15 elements. Judging from the plate lengths, the central opening was relatively small, at most one-third of the ring diameter. In Alrausuchus (Fig. 7b), the rings are preserved in both orbits of specimen PIN, no. 3706/10, although some elements are damaged or represented by imprints; however, a complete ring is reconstructed on each side. It is composed of almost flat elements in the shape of slightly elongated rectangles, with slightly elevated internal margin; as a result, the plates are slightly and gently sloping curve externally. The ring consists of 19 plates. The central opening is relatively small, its diameter is about 0.4 of the ring diameter. The formula of the ring is [+4–4+3–4], where (+) and (–) designate respective locking plates and figures are the numbers of intermediate plates, and the count starts from the anterodorsal positive lock upwards and backwards. In Titanophoneus (Fig. 7c), the ring is retained in the left orbit of specimen PIN, no. 157/1. All elements are completely or partially preserved. They are subquadrate, somewhat elongated, slightly curved on the internal side; the external surfaces are almost flat. The ring consists of 23 plates. The central opening is relatively large, at least 0.6 of the ring diameter. The ring formula is [+5–5+4–5] In Viatkogorgon (Fig. 7d), an almost complete right ring of specimen PIN, no. 2212/61 is preserved. The plates are narrow, almost quadrate; the external margin curves strongly and gently sloping medially; and the lower margin is elevated strongly laterally, so that the ring as a whole is almost truncated conical. The ring consists of 15 plates. The central opening is large, its diameter is at least 0.7 of the ring diameter. The ring formula is [+3–3+2–3]. The ring of this specimen was briefly described by Tatarinov (1999a); however, for uncertain reason, he indicated that the ring consists of two sectors. As the formulas of all known rings are compared, we see, that the odd number of plates is attributable to the posteroventral quadrant. The reasons for this remain uncertain. 2. Cheek Segment (ossa buccalia) The cheek, upper jaw segment is formed by the premaxillae (= ossa incisiva), septomaxillaries, and maxillaries. The premaxillae form the upper jaw symphysis, joining the symmetrical jaw segments. In Dicynodontida, these bones are densely fused up to complete oblitPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

887

eration of the suture. The vertical process of the premaxilla (processus ascendens, spina nasalis), which is articulated with the skull roof, usually overlaps considerably the nasal, approaching posteriorly the midlength of this bone (for example, in Ulemosaurus and Ulemica, see Ivakhnenko, 2003c, text-fig. 42). Only in Sauroctonus, Suchogorgon, and, probably, all Gorgonopidae, it does not reach the nasal (Ivakhnenko, 2005b, text-fig. 27). The labial plate of the bone overlaps the lateral surface of the maxillary, providing a rather firm articulation. However, in large predators (Biarmosuchus, Inostrancevia, Gorgonopidae), the surface of contact (sutura incisivomaxillaris) is gently sloping concave and almost smooth, of a mobile articular type (a sutural articulation is not developed). The external margin of the ventral surface of the bone body has a row of tooth alveoli (alveoli dentum incisivorum). There are usually four or five alveoli, although the number is not constant due to small caudal alveoli. Primitive Anomodontia have four (Ulemica) or five (Suminia); however, as the teeth of the incisive region are reduced, only two caudal alveoli are retained in Endothiodon (Cluver and King, 1983, text-fig. 9), while, in the other groups, they completely disappear. The palatine plate borders anteriorly the choana. In some taxa, the anterior margin of the choana is very wide (Archaeosyodon, Syodon, Titanophoneus, Ulemosaurus, and Inostrancevia). In other cases, the anterior part of the choana is more or less narrowed (to the formation of a narrow fissure slightly expanding in the anterior part). This structure of the anterior part of the choana is probably connected with the sensory organs located in this region (such as, for example, Jacobson’s organ: Ivakhnenko, 2005b, p. 438). The lower and posterior margins of a small septomaxillary usually adjoins the maxillary; at the bend in these region, the bone borders a more or less developed foramen between the septomaxillary and maxillary (foramen septomaxillaris). Internally, a groove extends to this foramen from the apertura anterior ductus lacrimalis (Fig. 1a). The posterior margin of the bone ascends along the anterior margin of the maxillary and usually forms the facial plate (pars facialis), which wedges in between the nasal and maxillary. The anterior margin of the septomaxillary forms the posterior border of the nares. This margin of the bone usually has a superficial anteromedial notch, which sometimes passes into a lateral depression (impressio lateralis) (Suchogorgon: Ivakhnenko, 2005b, text-fig. 27, ils). Above the notch, the bone margin curves medially, forming a small horizontal plate, the processus intermedialis. The upper margin of the plate is concave from above. The foramen anterior extending from the groove of the ductus lacrimalis usually opens in this depression. The typical structure of this bone is observed in all the Gorgonopia investigated, except for Viatkogorgon, in which the pars facialis is short and the foramen sep-

888

IVAKHNENKO (b)

(a)

(c)

(d)

Fig. 8. Septomaxillaries, right, external view: (a) Viatkogorgon ivakhnenkoi Tatarinov, 1999, holotype PIN, no. 2212/61; (b) Nikkasaurus tatarinovi Ivachnenko, 2000, holotype PIN, no. 162/33; (c) Syodon efremovi (Orlov, 1940), holotype PIN, no. 157/2; and (d) Alrausuchus tagax (Ivachnenko, 1990), specimen PIN, no. 3586/6. Scale bars: (a) 0.5 cm, (b) 0.2 cm, and (c, d) 1 cm.

tomaxillaris is an ovate depression (Fig. 8a), the lower part of which contains a small true septomaxillary foramen, and a groove extending to the foramen anterior passes along its upper margin. The structure of the septomaxillary of the Estemmenosuchidae, in which the posterior border of the nares is very pachyostotic, is obscure. In Nikkasaurus, the facial plate is in the shape of a thin process, which bypasses a large septomaxillary foramen and does not come in between the nasal and maxillary (Fig. 8b). In Reiszia, which is closely related to the last genus, a septomaxillary or a trace of its attachment have not been found. The facial plate is well developed in all the Dinocephalia investigated, although it is reduced considerably in Archaeosyodon, Syodon (Fig. 8c), and Titanophoneus. In these taxa, the impressio lateralis, is poorly developed; however, the processus intermedialis is a wide horizontal plate with an extensive dorsal depression. On the contrary, in Alrausuchus, the processus intermedialis is a vertical plate, curving medially and having a somewhat concave anterior margin (Fig. 8d); this depression is connected to the groove extending from the septomaxillary canal. In Ulemica and Suminia, the facial plate is relatively large (Fig. 1b); in Idelesaurus, it is reduced considerably; and in Dicynodon, Delectosaurus (Fig. 9a), and

other late Dicynodontidae, it is virtually absent. In Anomodontia, the processus intermedialis is also wide and horizontal, but the impressio lateralis is very well developed, expanding onto the anterior margin (near the nares) of the maxillary. The simplest septomaxillary is observed in Microurania; this is a thin facial plate along the anterior margin of the premaxilla, with a small subtriangular processus intermedialis, curving laterally in the lower part (Fig. 6b). Even this superficial review clearly shows that the septomaxillary requires a special study. Possibly, it is directly relevant to the glandula lateralis nasi, an important structure of primitive land tetrapods that dampened the olfactory tract in the region of the nares (see Tatarinov, 1976, pp. 37–43). The basic part of the cheek sector is formed by the maxillary. It usually has a high subtriangular vertical plate (lamina facialis), which in actual fact forms almost the entire cheek. As indicated above, this bone is weakly attached to the bones underlying the anterior part of the roof. Therefore, it is easily separated from the skull. In some specimens (e.g., Suchogorgon golubevi, specimen PIN, no. 4548/1, and Inostrancevia latifrons, specimen PIN, no. 2356/32), the maxillaries are isolated, while the lower jaws remain in natural articulation. Along the entire upper margin of the interPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA) (a)

889

(b)

Smx

sh

Fig. 9. Maxillaries: (a) Delectosaurus arefjevi Kurkin, 2001, holotype PIN, no. 4644/1, left, external view; and (b) Dicynodontidae gen. indet., specimen PIN, no. 4549/3, right, inner view. Designations: (sh) sinus maxillaris (sinus Highmori) and (Smx) septomaxillary. Scale bar, 1 cm.

nal surface of the lamina facialis, i.e., the area of articulation with the nasal, prefrontal, lachrymal, and jugal, the bone freely lies on wide shelves and is only attached by connective tissue. Even in Dicynodontidae, with the firmest attachment of the bone, it is usually found isolated. The thickened lower margin of the maxillary (pars alveolaris) has a row of alveoli. In Nikkasaurus, this region is most primitive. It has 15 or 16 alveoli gradually decreasing in size posteriorly. Reiszia has 13 or 14 somewhat larger alveoli, the fourth anterior alveolus is enlarged. Niaftasuchus has 12 alveoli decreasing in size posteriorly, the fifth anterior alveolus is strongly enlarged. Ulemosaurus and Deuterosaurus have 8 or 9 alveoli, the largest of which is the anterior. Among tapinocephalids, the greatest number of maxillary alveoli is probably in Taurocephalus, i.e., 12 (after Boonstra, 1936). Above the alveoli of enlarged teeth, the dorsal surface of the pars alveolaris usually has distinct tubercles (juga alveolaris). In taxa with the enlarged teeth developing in canines, a high sheath (capsule) for the canine (bursa dentia canini) is formed. Inside the capsule, there is a large alveolus of the canine (alveolus dentis canini). The canine alveolus is usually positioned almost vertically, at about 80° to the dental plate of the maxillary. However, in Archaeosyodon, the capsule is inclined posteriorly (at approximately 70°), and in Syodon, this angle is about 60° (Ivakhnenko, 2003c, textPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

fig. 10b). The canine capsule forms a more or less developed expansion on the intracranial surface (paries medialis) of the facial plate of the maxillary, which is an excessively developed juga alveolaris. In taxa with particularly large, saber-shaped canines, the posterior border of the alveolus contains an additional cavity, an alveolus of the replacement canine (canine successorum). In the Gorgonopidae (Pravoslavlevia, Sauroctonus, Suchogorgon), the alveolus of the replacement canine is located close to the alveolus of the main canine and, hence, a wall between them is almost absent. In predatory dinocephals (Alrausuchus, Biarmosuchus), the alveolus of the replacement canine is displaced somewhat posteriorly, and the canine capsule as though bifurcates in the upper part on the medial side (Tchudinov, 1983, text-fig. 8a); the wall between alveoli is only absent in the lower part. Inostrancevia has two or three additional alveoli, the lower part of the intracranial surface of the expansion of the capsule is widened sharply medially; the additional alveoli, which are connected longitudinally with the alveolus of the main canine are positioned anteromedially, medially, and posteromedially to it (Fig. 57a). A similar expanded capsule is observed in Kamagorgon and Admetophoneus. In Microsyodon, Archaeosyodon, Syodon, Titanophoneus, Parabradysaurus, and Estemmenosuchus, having narrow and high capsules, which as though envelope the

890

IVAKHNENKO

canine root, additional alveoli of replacement canines have not been recorded.

the maxillary sinus, since the crista infralacrimalis of the maxillary is almost completely absent.

Some taxa (all Gorgonopidae, Inostrancevia, and Biarmosuchus) lack precanine alveoli in the maxillary; Microsyodon, Microurania, and Titanophoneus have very small precanine alveoli. Posterior to the canine alveolus, there is a row of cheek teeth alveoli, which are designated postcanines (alveoli dentum postcaninorum). The number of postcanine alveoli is usual 8–10; however, in Gorgonopidae and Inostrancevia, there are at most four or five, and they are inclined obliquely posteriorly. The posterior inclination of alveoli is also characteristic of strongly klinorhinal taxa, such as Rhopalodon, Phthinosaurus, and Parabradysaurus. In Estemmenosuchus, the inclined position of alveoli is masked by their small size relative to a very massive maxillary. An unusual feature of this genus is a great number of cheek teeth (more than 25); five or six anterior teeth extend along the ventral side of the canine capsule, bypassing medially the canine (Ivakhnenko, 2003c, text-figs. 37c, 38b).

The intracranial surface of the pars alveolaris forms the lateral wall of the choana. In all taxa with a large upper canine, the projection of the canine capsule divides the choana into two parts; the anterior part contains the lower canine (fossa dentia canini inferioris) and the posterior part is the true choana. Among the taxa investigated, only Ulemica has a rudimentary maxillary hard palate; the medial margin of its pars alveolaris expands and distinctly overhangs the choana. Posterior to the choana, the maxillary comes in contact with the palatine. The contact surface between the paries medialis and ala maxillaris of the palatine of the majority of taxa (Microsyodon, Archaeosyodon, Syodon, Titanophoneus, etc.) is usually formed by two to four uneven longitudinal subhorizontal crests. In the Gorgonopidae, the contact surface has 10–12 very even, thin, parallel plates positioned obliquely (Ivakhnenko, 2005b, text-fig. 26). Ivakhnenko (2005b, p. 439) proposed that the unique suture was connected with the kinetic maxillary. In Inostrancevia, the suture is even more simplified, having two or three low oblique parallel plates, which are positioned even closer to the vertical than in Gorgonopidae. In Dicynodon, Delectosaurus, and related taxa, the pterygoid reaches the maxillary, overlying externally a thin projection of its margin, and the palatine overlies this projection medially, forming a dense suture that provides firm attachment of the maxillary to the palatal segment.

Primitive Anomodontia (Otsheria, Ulemica, Suminia) have 8–10 cheek alveoli, which gradually decrease in diameter posteriorly. In Ulemica, the fourth anterior alveolus is enlarged. Endothyodon retains seven alveoli of the cheek teeth (Cluver and King, 1983, text-fig. 9). The majority of higher Dicynodontida have a strongly increased canine-shaped tooth (tusk) and a more or less increased alveolus. Dicynodonts lack a capsule for the tusk; thus, breaking through the roof of the juga alveolaris, it enters the sinus Highmori and, in the case of considerable development, fills it almost completely (Fig. 9b). Only a thin plate bordering the tusk cavity anterior to the olfactory– narial region usually develops. The first maxillary alveolus is increased and usually followed by several small alveoli (up to six in Pristerodon: Cluver and King, 1983, text-fig. 12b; one or two in Eodicynodon and Tropidostoma: Cluver and King, 1983, text-figs. 7, 14). Australobarbarus has two or three alveoli of small teeth on a small tubercle posterior to and somewhat internal to the tusk alveolus. In the majority of taxa (Idelesaurus, Dicynodon, Vivaxosaurus, Delectosaurus, Elph, etc.), the alveolus for the tusk is only retained; however, it sometimes disappears (for example, in the Triassic Placeriidae). The dorsal surface of the pars alveolaris forms the bottom of the maxillary sinus (antrum of Highmore, sh). In canineless taxa (Otsheria, Ulemica, Suminia, Fig. 1a), the cavity is separated dorsally by a crest (crista lacrimalis) from the upper olfactory capsule and by the infralacrimal crests (located on the lachrymal and maxillary) from the lower olfactory capsule. In taxa with a high canine capsule (Biarmosuchus, Titanophoneus, Syodon, Inostrancevia, and Suchogorgon: Ivakhnenko, 2005b, text-fig. 26), the upper cavity is extremely poorly developed, while the lower cavity is fused with

The posterior margin of the maxillary is shaped like a more or less sulcate process underlying the jugal. In Dicynodon, Vivaxosaurus, Delectosaurus, and Diictodon (Cluver and King, 1983, text-fig. 30), this process extends far posteriorly along the lower external surface of the zygomatic arch, replacing the jugal on this surface and reaching the squamosal. The external surface of the facial plate of the maxillary (facies facialis) is usually covered with surface sculpture. In a number of taxa (Titanophoneus, Ulemosaurus, Inostrancevia, and all Dicynodontida), this surface is almost smooth and has a few relatively large foramina for blood vessels, giving rise to the grooves extending on the surface (Fig. 9a). In the central part of the plate, the grooves curve, sometimes anastomose with each other, becoming almost straight at the bone margins. In taxa with thickened bones (Alrausuchus, Biarmosuchus, Kamagorgon, Suchogorgon, Sauroctonus), the facial surface is covered with a background sculpture, like the bones of the skull roof. This sculpture is formed of small tubercles, which are sometimes fused in short curving ridges, with small, widely spaced foramina for blood vessels between them. The ridges usually become elongated towards the bone margins. On the margin of the facial lamina, mostly along the sutura incisivomaxillaris and particularly along alveolar margin, the bone surface is relatively smooth, with large widely spaced foramina. Short grooves usually PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

(b)

891

(a)

Fig. 10. Left maxillaries, external view: (a) Biarmosuchus tener Tchudinov, 1960, specimen PIN, no. 1758/85; and (b) Archaeosyodon praeventor Tchudinov, 1960, specimen PIN, no. 1758/118. Scale bar, 1 cm.

extend from the foramina to the bone margin, expanding or diverging distally. In the taxa listed and also Microurania, in the central part of the facial lamina, the background sculpture is supplemented by circular pits, which are almost equal in size (0.6–1 mm) in all taxa (even in very small Microurania). The bottom and walls of the pits are smooth. They occur in almost all sites of the surface, partially overlying the foramina and grooves of the background sculpture. The pits are located at a distance from each other or their edges are almost fused. They are scattered randomly, without forming anything like rows (Fig. 10a). The field of pits is shaped like an irregular vertically extended oval. The surface sculpture of the facial lamina of Microsyodon, Archaeosyodon, and Syodon is a usual background sculpture, distinguished by the very small tubercles and ridges looking like shagreen surface (Fig. 10b). The foramina between tubercles are also small and numerous; thus, the bone surface is fineporous. In Estemmenosuchus, the facial lamina is covered with rugose expansions and tubercles, with a background sculpture on their surfaces. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

3. Temporal Segment (ossa temporalia) Bones of the temporal segment (jugals, postorbitals, and squamosals) border posteriorly and dorsally the orbit and the synapsid temporal fenestra and, hence, form the zygomatic arch. The anterior margin of the jugal (= zygomatic) participates in the formation of the cheek segment. The pars facialis of this margin forms a part of the facial plane and, along with the horizontal plate (lamina ventralis) and transverse plate (lamina transversa), forms the posterior part of the lower olfactory capsule (Fig. 1a), which is fused topographically with the maxillary sinus in taxa with a large canine capsule (Ivakhnenko, 2005b, text-figs. 22, 26). In a number of taxa (Alrausuchus, Biarmosuchus, Suchogorgon, Sauroctonus, Pravoslavlevia, Viatkogorgon, and Inostrancevia), the facial plane of the pars facialis has a flat depression (impressio antorbitalis). This depression is widely open anteriorly towards the region of the maxillojugulare suture. It is often covered with semicircular rugose crests, the convex side of which is turned posteriorly. Between crests, there are several large foramina, varying in position and sometimes passing into branching grooves. In Inostrancevia, the impressio antorbit-

892

IVAKHNENKO

alis is deepened and occupies a large area, passing onto the entire posteroventral margin of the facial surface of the lamina facialis of the maxillary. In some taxa, such as Archaeosyodon, Syodon, and Titanophoneus, the depression is also recognizable, although it is longitudinally elongated, flattened, superficial, slightly expands onto the posterior margin of the maxillary. Ivakhnenko (2005b, p. 441) proposed that, in taxa with a possibly somewhat mobile maxillary, this was the attachment area for a muscle involved in this mobility. In all known Anomodontia with a clearly immobile maxillary, its posterior margin closely approaches the orbit, and a similar structure is absent. Farther posteriorly, the jugal forms an expanded infraorbital border of the orbit (margo infraorbitalis) and has a contact area for the postorbital. The posterior margin of the bone (processus zygomaticus) forms almost the entire lower margin of the zygomatic arch (arcus zygomaticus). In primitive taxa, this process is thin, positioned almost horizontally (Niaftasuchus) or even slightly elevated at the base (Microurania, Nikkasaurus) and curves considerably downwards in all Gorgodontia. In almost all taxa investigated, even those without pachyostotic expansions in other areas of the skull, the lower margin of the process is thickened and frequently has pachyostotic expansions, covering dorsally the region of the recessus alae angularis. The most expanded and lowest lower margin of the zygomatic process is in Syodon, Titanophoneus, Deuterosaurus, and Ulemosaurus (Ivakhnenko, 2003c, text-figs. 12, 13, 18, 19), all Phthinosuchidae (Ivakhnenko, 2003c, textfig. 27b), Rubidgeidae (see Sigogneau, 1970a), and Estemmenosuchus, in which the pachyostotic expansion of the jugal is relatively small in young animals, becoming a very long, subhorizontal process in adults (Ivakhnenko, 2003c, text-figs. 37b, 38). In the higher Dicynodontidae (Dicynodon, Delectosaurus), the zygomatic process of the jugal underlies a long process of the maxillary; hence, the jugal forms only the internal margin of the zygomatic arch. Anomodontia are characterized by an elevated zygomatic arch, with the subapsid fenestra (the functioning structure of the temporal region), the anterior margin of which comes under the zygomatic process of the jugal. In primitive Ulemicia (Otsheria, Ulemica), the zygomatic process of the jugal is elevated slightly above the lower border of the orbit (as in Eodicynodon: Rubidge, 1984). In Australobarbarus (and Pristerodon: Cluver and King, 1983, text-fig. 13c), it is approximately at the level of the orbital center; in Idelesaurus (and Oudenodon: King, 1988, text-fig. 21c), it is positioned slightly higher; and in all Dicynodontidae (Dicynodon, Delectosaurus, etc.), it is above the upper orbital border. The postorbital plays a significant role in the formation of the temporal segment. This bone partially forms the postorbital region of the skull roof, and, sharply curving downwards, borders the postorbital arch and forms the caudal process (processus caudalis), overlapping from above the external margin of the temporal fenestra.

The postorbital arch of the majority of taxa is thin (Gorgonopia, Anomodontia); its intracranial surface has a low subvertical crest (crista postorbitalis), which corresponds to the linea temporalis anterior, delineating the anterior margin of the temporal fossa. This crest usually passes along the posterior margin of the bone. Only in the Phthinosuchidae, Rubidgeidae, Inostrancevia, and Estemmenosuchus (Ivakhnenko, 2003c, textfigs. 27a, 27b, 30, 36a, 37b), the postorbital arch sharply expands in ontogeny and covers externally the anterior part of the temporal fossa; thus, the crest is displaced anteriorly along the internal (intracranial) surface. An unusual structure of the postorbital arch is characteristic of Dinocephalia. In the most primitive taxa (e.g., Niaftasuchus, Fig. 11a), a gentle depression (fovea supraorbitalis) is formed at the bend of the postorbital between the planes of the skull roof and the postorbital arch. In this area, the origin of muscles is displaced on the external side and, hence, the postorbital crest passes onto the dorsal surface anteriorly from the posterior margin of the postorbital. The depression formed in this area initially occupies a part of the postorbital (Alrausuchus, Fig. 11b) and, then, expands anteromedially, reaching the suture between the postorbital and postfrontal (Archaeosyodon, Fig. 11c); further, becoming deeper, it expands onto the dorsal surface of the postfrontal (Syodon, Titanophoneus, Fig. 11d); in the case of extreme development, it reaches the frontal (Deuterosaurus: Ivakhnenko, 2003c, text-fig. 18b; Ulemosaurus). The caudal process of the postorbital forms the dorsoexternal wall (planum temporale) of the adductor fossa. The upper surface of the caudal process is smooth, almost a half of its medial surface is overlain by the facies postorbitalis of the parietal, although a sutural surface (facies parietalis) is not formed. The absence of sutural contact in this place is probably connected with the preservation of a rudimentary parapsid fissure, which opens in parapsids and diapsids as the upper temporal fenestra (see Ivakhnenko, 2005b, p. 436). The caudal process comes in contact with the anterodorsal margin of the squamosal, and its lower margin enters a groove (sulcus postorbitalis ossis squamosi) on the supramedial margin of the ala temporalis of the squamosal. With reference to the postorbital arch, the caudal process is initially positioned subhorizontally, sometimes slightly elevated; this increases the area of the temporal fenestra. However, in taxa with a pronounced klinorhiny (Syodon, Archaeosyodon, Ulemosaurus, and Estemmenosuchus), the caudal process is directed somewhat downwards. The temporal fossa is bordered posterolaterally by the squamosal. In fact, this bone is a lateral cover of the palatoquadrate region that is homologous to the traditional squamosum (= paraquadratum: Gaupp, 1906) of, for example, lizards (Ivakhnenko, 2003c, p. 356; 2005b, p. 435) rather than to the squamosum (= os temporale) of mammals. This bone is exposed to several PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA) (a)

893

(b)

pc

Pof

cppo cppo

Pof cppo

cppo (c)

(d)

pc

P

pc

Fig. 11. Left postorbitals: (a) Niaftasuchus zekkeli Ivachnenko, 1990, specimen PIN, no. 162/63, juvenile, lateral view; (b) Alrausuchus tagax (Ivachnenko, 1990), holotype PIN, no. 3706/10, lateral view; (c) Archaeosyodon praeventor Tchudinov, 1960, specimen PIN, no. 1758/93, lateral view; and (d) Titanophoneus potens Efremov, 1938, lectotype PIN, no. 157/1, dorsal view. Designations: (cppo) crista postorbitalis of the postorbital, (P) parietal, (pc) caudal process of the postorbital, and (Pof) postfrontal. Scale bars: (a, b) 0.5 cm and (c, d) 1 cm.

loads; therefore, it is complex and variable, although almost uniform in general design in all taxa investigated. The thin vertical plate of the bone body borders posteriorly the temporal fossa and its medial margin adjoins the parietal region and the periotic. As mentioned above, the articulation with the parietal shield and periotic is mediated by the supratemporal (if it is preserved); as the supratemporal is reduced, the upper margin of the squamosal adjoins anteriorly (that is, PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

topographically from below) the angulus parietalis without forming a distinct suture. In the absence of the supratemporal, on periotic contact with the supraoccipital process of the synotic and paroccipital process of the opisthotic, the surface of the squamosal is smooth and somewhat lowered because of the preservation of rudimentary primary fissure (hiatus paraoticalis) between the structures of the periotic and the squamosal, which covers laterally cartilage bones of the palato-

894

IVAKHNENKO rts

(b)

P

(a) F

Pp

fep

alos T

pd

fal

Sq

Sq

S pb Fig. 12. The occipital region of Biarmosuchus tener Tchudinov, 1960, reconstruction based on holotype PIN, no. 1758/2 and specimens PIN, nos. 1758/19 and 85: (a) anterior and (b) posterior view. Designations: (alos) left ala orbitosphenoidei, (F) frontal, (fal) facies laterosphenoidalis of the prootic, (fep) facies epipterygoidei of the parietal, (P) parietal, (pb) basipterygoid process of the basisphenoid, (pd) dorsal process of the prootic, (pmt) pars mastoideus of the tabular, (Pp) postparietal, (rts) recessus tecti synotici of the parietal, (S) stapes, (Sq) squamosal, and (T) tabular. Scale bar, 1 cm.

quadrate region. Since jaw muscles adjoin anteriorly this weakened region, it is overlain by a thin superficial plate (ala perioticalis), which varies in the extent of development in different taxa. In Dinocephalia and Anomodontia, this plate is relatively small in area (Figs. 12a, alos; 13a); its smooth internal surface covers anteriorly the lower margins of the parietal, postparietal, and margins of the periotic. In Gorgonopidae, the ala perioticalis is divided into two rami, the upper pars parietalis and the lower pars synoticus, which extend somewhat farther along the periotic than in other groups (Ivakhnenko, 2005b, text-fig. 8). The dorsolateral margin of the squamosal, which borders posteroventrally the temporal fossa, is almost vertical in Dinocephalia. In primitive Anomodontia (Ulemica), it curves slightly posteriorly; in Suminia, its posterior margin curves somewhat ventrally (Fig. 13b). The same curvature is slightly developed in Dicynodontida. In Gorgonopida, the curvature and posterior expansion of this region of the squamosal become func-

tionally important. In this group, the dorsolateral margin (ala temporalis) forms the expanded upper part of the temporal fossa, significantly increasing the volume of the upper part of the cavity (Ivakhnenko, 2005b, p. 419). After the bend of the plate of the ala temporalis, its anterior surface, which initially faces the temporal fossa, turns dorsally. As a result, the posterolateral margin of the squamosal passes posteriorly much farther than the planes of the occipital edge. This is particularly well pronounced in Inostrancevia (Ivakhnenko, 2003c, text-fig. 29b), while, in the Rubidgeidae, due to the strong development of the ala temporalis and expansion of the postorbital arch, the plane of the temporal fenestra acquires a somewhat dorsolateral orientation (Sigogneau, 1970a: Dinogorgon, pl. 75b; Rubidgea, pl. 79; Broomicephalus, pl. 80). The anteroventral margin of the squamosal has a concave area for attachment of the QQJ-complex. In primitive taxa that have been examined with reference to the structure of this region (Alrausuchus) and BiarPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA) ffP

(a)

895

(b)

fPo

fPf Sq

pmt fel

QQJ S

tv

fPh

pb

Fig. 13. The occipital region of Suminia getmanovi Ivachnenko, 1994, reconstruction based on holotype PIN, no. 2212/10 and specimen PIN, no. 2212/62: (a) anterior and (b) posterior views. Designations: (fel) fossa elliptica of the squamosal, (ffp) facies frontalis of the parietal, (fpf) facies postfrontalis of the parietal, (fph) facies rostri parasphenoidei of the basispheniod, (fpo) facies postorbitalis of the parietal, (pb) basipterygoid process of the basisphenoid, (pmt) pars mastoideus of the tabular, (QQJ) quadrate–quadratojugal complex, (S) stapes, (Sq) squamosal, and (tv) tuba vestibuli. Scale bar, 1 cm.

mosuchus, Suchogorgon, and Inostrancevia, which retain the streptostylic QQJ-complex, this area (fossa quadratica) is deeply concave, with a concave and expanded upper margin (crista superior), and the capitulum quadrati enters the fossa quadratica with a significant gap (for detailed description of Suchogorgon, see Ivakhnenko, 2005b, p. 419). In this design, the plane of the quadratojugal is positioned almost perpendicular to the plane of the quadrate, and it moves relative to the flat internal medial surface of the base of the zygomatic process of the squamosal. In all other Dinocephalia and all Anomodontia (from the level of Ulemicida), with an immobile QQJ-complex, the articular area expands and becomes less concave, a gap between it and the caudal surface of the complex disappears, and the quadratojugal approaches the orientation of the quadrate, and tightly adjoins the flat lateral part of the expanded articular area. In Dicynodon, Delectosaurus, Vivaxosaurus, and others, the articular area of the wide quadratojugal is even separated by a low crest from the area of the quadrate. In the groups with a low zygomatic arch (Gorgonopia, Dinocephalia), the zygomatic process of the squamosal deviates at the level of the articular area of the PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

QQJ-complex. It forms the posterior part of the zygomatic arch, completely bordering from above a very small subapsid notch. The contact area for the jugal is usually narrow; however, in Ulemosaurus, in connection with the expansion of the zygomatic arch, the zygomatic process of the squamosal expands sharply anteriorly (Ivakhnenko, 2003c, text-fig. 19a). In Anomodontia, in connection with the strong development of the subapsid fenestra, the base of the zygomatic process is positioned much higher than the fossa quadratica. The bone segment between the base of the zygomatic process and the fossa quadratica of Ulemica is in the shape of an ovate depression open mostly anteriorly (in the temporal fossa) and slightly laterally (Ivakhnenko, 2003c, text-fig. 42). In Suminia (Ivakhnenko, 2003c, text-fig. 43), the upper and lateral margins of the depression curve externally; in the higher Dicynodontida, the entire lateral margin of the squamosal in the region of the subapsid incisure curves externally; thus, the anterior surface of the bone in the region of the fossa quadratica is obliquely laterally oriented, and the region of the base of the zygomatic process is oriented almost laterally.

896

IVAKHNENKO (a)

Pp

(b) T

fel

fel

Fig. 14. Occipital region, posterior view: (a) Ulemosaurus svijagensis Riabinin, 1938, lectotype PIN, no. 2207/2; and (b) Delectosaurus arefjevi Kurkin, 2001, holotype PIN, no. 4644/1. Designations: (fel) fossa elliptica of the squamosal, (Pp) postparietal, and (T) tabular. Scale bars: (a) 5 cm and (b) 1 cm.

Two structures on the occipital surface of the squamosal deserve consideration. The bone, along with the pars mastoideus of the tabular (and crista mastoidea of the supratemporal, if it is present), form a massive processus mastoideus (Fig. 12b, pmt). On its surface, there were origins of the aponeuroses of the sternocleidomastoid muscle, which extended from the region of the sternoclavicular articulation. In all the Anomodontia examined, the mastoid process is very poorly developed (Fig. 13b).

tal flange of the squamosal is widened throughout its extent and the caudal surface is concave at the level of the fossa quadratica, forming the fossa elliptica. This is a very small depression, which is poorly pronounced, for example, in Ulemica or Suminia (Fig. 13b), but is elongated and passes far upwards, approaching the upper margin of the squamosal, for example, in Ulemosaurus and Deuterosaurus. The lower margin of the squamosal tightly adjoins the body of the quadrate.

The second structure of the occipital surface of the squamosal is only present in some taxa. In Alrausuchus (Fig. 39b), Biarmosuchus, and Gorgonopidae (Ivakhnenko, 2005b, text-fig. 7), the squamosal comes onto the occipital surface as a narrow flange (crista posterior), which expands somewhat and is convex in line with the fossa quadratica. The medial margin of this flange slightly deviates from the body of the quadrate in the lower part. In some other taxa, such as Syodon, Archaeosyodon (Fig. 15b), Ulemosaurus (Fig. 14a), Estemmenosuchus (Fig. 15a), and Delectosaurus (Fig. 14b), the crista posterior is absent, but the occipi-

4. Palatal Segment (ossa palatalia) The bones of this segment underlie the ethmoid region, participate in the division of air-conducting and mouth cavities, connect the cheek segment with the braincase. This region is formed by the vomer, palatines, pterygoids, ectopterygoids, and parasphenoid. The epipterygoids and quadrate–quadratojugal complex should also be considered in this section. The support of the ethmoid region involves the vomer, palatines, and pterygoids. The posterior part of the dorsal surface of the vomer passes into a thin vertiPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA) (a)

897

(b) T

Sq T poc

fel

poc

fel

Qj

Pt Fig. 15. Part of the occipital region, posterior view: (a) Estemmenosuchus mirabilis Tchudinov, 1968, holotype PIN, no. 1758/6; and (b) Archaeosyodon praeventor Tchudinov, 1960, specimen PIN, no. 1758/93. Designations: (fel) fossa elliptica of the squamosal, (Pt) pterygoid, (Qj) quadratojugal, (poc) paroccipital process, (Sq) squamosal, and (T) tabular. Scale bar, 2 cm.

cal plate, which bifurcates upwards and forms the alae vomeris, which cover the septum nasi osseum and the anterior margins of the ethmoid septum. Similar plates (alae parasagittalis) are observed on the dorsal surface of each palatine and pterygoid. The plate of the palatine ascends higher than the plate of the vomer, partially covering the posterodorsal margin of the latter; hence, the plate of the pterygoid is also located higher than the plate of the palatine and overlies its upper margin. At the medial line, the plates of opposite bones adjoin each other, slightly diverging at the upper margin and covering from below a part of the lateral surface of the rostrum parasphenoidale, the anteroventral part of the presphenoid, and the lower part of the mesethmoid, reaching the septum nasi. In fact, the anterior margin of the palatal region is formed by the palatine plate of the premaxilla, which along with the maxillary border the anterior and lateral margins of the pars anterius (caninus) of the bony choana. Medially, the choana is bordered by the vomer. In Nikkasaurus, the bone is paired and has a row of small teeth along the medial suture. In the majority of taxa, the bones are fused, although certain traces of the medial suture are usually observed on the posterior margin of the ventral surface. In the posterior part of the bone, in the region of the pars posterius of the bony choana (air-conducting part, naria interna), PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

the lateral margins of the bone are thickened, curve downwards (ala ventralis), approach each other, and, on the ventral surface of the bone, come in contact to form a triangular tubercle (eminentia interchoanalis). In Gorgonopidae, a narrow posterior end of the vomer enters between the palatines (Ivakhnenko, 2005b, text-fig. 17). In the other taxa examined, the posterior end of the vomer expands, spreads under the palatines, and a vertical plate wedges in between them, bifurcates, and forms the alae vomeris. Only very primitive taxa (Nikkasauridae, Niaftasuchus, Alrausuchus, Ulemicida) have an almost horizontal lower margin of the vomer. In all higher Gorgonopia and Dinocephalia, it curves distinctly upwards, forming a high nasopharyngeal meatus above the region of the naria interna. In Estemmenosuchus, the plate of the vomer curves only slightly; however, the area of the ala ventralis and eminentia interchoanalis expands laterally; in Estemmenosuchus uralensis, it is flattened and equipped with small teeth. The nasopharyngeal meatus of these taxa is separated from the mouth cavity by expansions on the ala ventralis. The caudal margin of the pars posterius of the bony choana is bordered by the palatine. In almost all taxa, the choanal incisure in the anterior margin of the bone has a gently sloping anterior elevation, which continues

898

IVAKHNENKO (a)

ppc Pt

Pl (b)

(c) Ect

Fig. 16. Structural elements of the palatal region: (a) Biarmosuchus tener Tchudinov, 1960, palatal surface, ventral view, reconstruction based on specimens PIN, nos. 1758/18 and 255; (b) Syodon efremovi (Orlov, 1940), specimen PIN, no. 157/635, right palate, medial view; and (c) Titanophoneus potens Efremov, 1938, specimen PIN, no. 157/186, tooth of the pterygoid flange. Designations: (Ect) ectopterygoid, (Pl) palatine, (ppc) pars posterius of the bony choana, and (Pt) pterygoid. Scale bars: (a, b) 1 cm and (c) 0.2 cm.

the nasopharyngeal meatus. Posteriorly, this meatus in many taxa (Niaftasuchus, Alrausuchus, Biarmosuchus, Archaeosyodon, Syodon, Titanophoneus, Dinosaurus, Kamagorgon, Proburnetia, Estemmenosuchus, and all Gorgonopidae) is bordered by a laterally elongated or V-shaped ridge (tuberculum palati), which usually has small teeth and forms the lateral walls of the nasopharyngeal meatus (Fig. 16a). In Ulemosaurus and Deuterosaurus, the tubercles are circular, have from one to three teeth on the apex (Ivakhnenko, 2003c, textfigs. 18c, 19b). The tubercles on the palatines are connected by more or less developed crests to tubercles on the pterygoids. The tubercle on the pterygoid is more variable in shape than the palatine tubercle. This tubercle is usually narrow V-shaped, with a row of small teeth on each branch (Nikkasaurus, Reiszia: Ivakhnenko, 2000b, text-fig. 3b); wide V-shaped, with many small teeth (Biarmosuchus, Archaeosyodon, Dinosaurus, Kamagorgon, Viatkogorgon, Estemmenosuchus: Ivakhnenko, 2003c, text-figs. 4c, 9c, 25c, 26b, 27c, 37c), with a few larger teeth (Suchogorgon: Ivakhnenko, 2005c, text-fig. 5b), or without teeth (Sauroctonus); small circular, with two or three large teeth at the apex (Niaftasuchus: Ivakhnenko, 2003c, text-fig. 1b), or almost completely reduced with one or two small teeth (Syodon, Titanophoneus: Ivakhnenko, 2003c, textfigs. 12c, 13c). These variations are probably attribut-

able to the fact that the pterygoid tubercle is virtually excluded from the alimentary function, and, if present, only restricts the air passage. In Inostrancevia (Ivakhnenko, 2003c, text-fig. 29c), the ventral surface of the palatines is completely smooth, lack a trace of palatine teeth or tubercles. In primitive Anomodontia, the palate in this region shows almost the same structure as in the groups considered above (Ulemica even retains low palatine crests in place of tubercles: Ivakhnenko, 2003c, text-fig. 42c). The palatine and ectopterygoid participate in the attachment of the lateral margin of the palatine plate with the bones of the cheek and temporal regions. The variants of attachment of the palatine to the maxillary are described above. A small surface plate of the ectopterygoid is usually rounded rhombic or longitudinally extended (Biarmosuchus). The posterolateral margin of the bone, along with the crista transversalis of the pterygoid, adjoin the ala ventralis of the jugal, and the posterior margin adjoins the anterolateral margin of the crista transversalis of the pterygoid. The sutures between the ectopterygoid, maxillary, and palatine usually has a small foramen (foramen palato-nasalis) providing passage for the lateral branches of the palatine artery and palatine nerve (see Tatarinov, 1976, pp. 114, 115). In Suchogorgon, the canal passes through the ectopterygoid (Ivakhnenko, 2005b, p. 422). PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

In the majority of groups, the palatine plate is separated caudally from the basitemporal region by the line of the pterygoid flanges (cristae transversalia). The flange is formed due to the ventral curvature of the posterior margin of the pterygoid. The lower margin of the crista transversalis is thickened medially, with an expanded surface, which forms an elongated area fovea denticulata. This surface occasionally has a row of teeth, varying in number and size in different taxa. In large Dinocephalia (Titanophoneus, Ulemosaurus, Deuterosaurus), the pterygoid flanges are thickened, very massive, and retain two or three relatively small, irregularly arranged teeth. The lateral margin of the flange (apex cristae transversalis) is widened and forms a surface (facies mandibularis), which was probably covered by a cartilaginous layer and provided shock absorption in the case of lateral stresses on the jaw. Within Anomodontia, Ulemicida (Ivakhnenko, 2003c, text-figs. 40b, 41b, 42c) and Suminia (Ivakhnenko, 2003c, text-fig. 43b) and Eodicynodon (Cluver and King, 1983, text-figs. 3A, 3B) retain small projections of the pterygoid flange. Like Suchogorgon (Ivakhnenko, 2005b, p. 438), all taxa with well-developed cristae transversalia lack a trace of attachment of the anterior pterygoid muscles on the posterior surface of the cristae, and, probably, in all cases, these muscles were relatively poorly developed. The greatest constructional changes in the palatal segment are observed in the higher Dicynodontida in connection with the development of the palatine plate of the premaxilla. In the primitive case, the anterior margin of the bony choana was located anterior to the line of the maxillary (Ulemicida, Suminia); subsequently, it was displaced close to the line of the midlength of the maxillary (Australobarbarus, Eodicynodon, Endothiodon: Cluver and King, 1983). In the higher Dicynodontida, this margin is approximately in line with the posterior margin of the maxillary or even posterior to it (for example, in Dicynodontidae and others). Thus, the palatal plates of the palatine and pterygoid are as though displaced posteriorly, they narrow and curve somewhat dorsally, forming the lateral border of the nasopharyngeal canal. Regarding the isolation of the air-conducting and alimentary systems, certain unusual formations in this region of Suminia, Deuterosaurus, and Biarmosuchus are of interest. The palate of Suminia (specimen PIN, no. 2212/62) has thin bone plates with many small teeth somewhat anterior to the region of the basipterygoid articulation (Fig. 17a). The plates are formed of coarsecellular tissue; the teeth are narrow, pointed, slightly curved. As the skull of Deuterosaurus biarmicus (specimen PIN, no. 1954/1) underwent chemical treatment, the mouth cavity showed several small teeth (Kurkin, 1997), which, judging from their size, were neither jaw nor palatal teeth (Fig. 17b). These teeth have long roots and complex crowns sharply different from the simple conical crowns of palatal teeth of Dinocephalia (comPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

899

pare Figs. 17b, 17c). In Biarmosuchus tener (specimen PIN, no. 1758/255), chemical treatment of the anterior skull part has revealed that the lateral surfaces of the vertical plate of the vomer are covered with fields of very small teeth. The teeth (dentis palatalis accessoria) are very thin-walled, pointed conical, extend obliquely posteriorly in poorly pronounced rows, with a disordered arrangement of teeth in rows (Fig. 17c). The tooth bases are connected to the bone surface by rough vesicular bony tissue. The tooth apices are pointed, curved downwards. The largest teeth are positioned relatively close to the sutures with the palatines, becoming somewhat smaller anteriorly. It is interesting that relatively narrow rows of similar small teeth even spread onto the lateral surfaces of the internasal septum. An isolated find of these formations could have been attributed to an admixture of foreign objects (although it should be noted that, in all cases, no foreign bones have been recorded in the localities). Three finds of this sort suggest that the air-conducting passage and the mouth cavity of at least some taxa were isolated by a soft palate underlying the pars anterius of the bony choana or even the entire nasopharyngeal meatus. The compacted borders of openings in the soft palate could have been covered with tooth plates or isolated teeth, which intercepted food from air passage. To date, similar bone plates with shagreen teeth in the palatal region have only been described in various labyrinthodont groups (for review, see Schoch, 2006). On the contrary, the presence of skin teeth on the vomerine surface of Biarmosuchus would rather be treated as a sign of the absence of a soft palate. A very important function of the pterygoid is attachment of the palatine plate to the braincase, which is performed by a flat caudal process (processus basicranialis). The basicranial process is developed to a varying extent in different taxa, its wide base deviates from the corpus pterygoidei at the point of curvature to the crista transversalis; medially, the process comes in contact with the braincase (in primitive tetrapods, with the membrane parasphenoid bone; as it is reduced, with the area between the epipterygoid and the basisphenoid process). In a primitive design, a narrow fissure (incisura interpterygoidea), which is retained in the majority of primitive taxa (Nikkasaurus, Reiszia, Ulemica, Otsheria, and probably Niaftasuchus), is located between the basicranial processes. A relatively small incisura interpterygoidea is retained in all Anomodontia, in which the margins of the pterygoids in this region deviate from the sharply dorsally curved rostrum parasphenoidale, and the alae parasagittales deviate dorsally from the level of the anterior margin of the incisura interpterygoidea. In representatives of all other groups under study (Dinocephalia, Gorgonopia), the medial margins of the pterygoids adjoin the rostrum parasphenoidale. If this region retains on the ventral side a longitudinal depression corresponding to the incisura interpterygoidea (Niuksenitia, Proburnetia, etc.), it is overlain by the alae parasagittalis. In all Gorgonopidae

900

IVAKHNENKO (a)

(b) Pt

pb (c)

Bs

fpv

Fig. 17. Palatal structures and palatal teeth: (a) Suminia getmanovi Ivachnenko, 1994, specimen PIN, no. 2212/62, part of the palatal surface; (b) Deuterosaurus biarmicus Eichwald, 1846, holotype PIN, no. 1954/1, dents palatalis accessoris; and (c) Biarmosuchus tener Tchudinov, 1960, specimen PIN, no. 1758/255, vomer, left view. Designations: (Bs) basisphenoid, (fpv) facies palatini of the vomer, (pb) basipterygoid process of the basisphenoid, and (Pt) pterygoid. Scale bars: (a, b) 0.1 cm and (c) 0.5 cm.

and Dinocephalia, the margins of the pterygoid not only adjoin the rostrum parasphenoidale throughout its length, but also have a more or less developed thin plate (lamina basicranialis), which is displaced ventrally and borders the rostrum. The length of the processus basicranialis varies widely. In Nikkasaurus, this region is the shortest; it is also poorly developed in all Anomodontia. The basicranial processes of Dinocephalia and Gorgonopia are well developed. The length of the basicranial processes is clearly connected with the elongation of the fossa temporalis and the length of the braincase. The most developed basicranial processes are in taxa with a relatively short braincase and elongated fossa temporalis (Gorgonopidae, Inostrancevia, Biarmosuchus, Ulemosaurus), in which the region of the basipterygoid articulation is displaced in the posterior part of the fossa temporalis. In Dicynodontia, this region is displaced far posteriorly, although the basicranial processes are rela-

tively short and the displacement results from the posterior lengthening of the entire palatal plate. The caudal margin of the ventral surface of the basicranial process has a groove (sulcus epipterygoideus), which contained the base of the epipterygoid. The medial margin of the groove is usually covered dorsally by a greater or lesser developed projection of the medial wall (lamina accessoria). In this region, a flat process (processus quadrati) deviates posteriorly and laterally from the processus basicranialis and extends to the quadratum, bordering medially the pterygoid incisure (fissura pterygoidea), the medial margin of the middle part of the temporal fossa. In Suchogorgon (Ivakhnenko, 2005b, text-fig. 23), the process is very short and forms a pointed descending hamulus pterygoideus. In Alrausuchus, Biarmosuchus, and Inostrancevia, the processus quadrati is longer, although it does not reach the quadrate. In the other taxa examined in regard to this region (Archaeosyodon, Syodon, Titanophoneus, PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

Deuterosaurus, Ulemosaurus, Estemmenosuchus, Ulemica, Suminia, Delectosaurus), the posterior margin of the processus quadrati is widened and forms dense angular contact with the quadrate, leaning anteriorly on its lateral margin and forming a thin plate that bypasses laterally. Formally, the palatal segment is terminated by the parasphenoid, which spreads under the braincase. In tetrapods, this bone usually develops from several centers of ossification. At least in Dinomorpha, its anterior region (rostrum parasphenoidale) and posterior wings (alae parasphenoidalis) distinctly show independent formation. In Biarmosuchus tener (specimen PIN, no. 1758/85), the rostrum is detached, although it is not broken off (Fig. 32d, fph); the parasphenoid wings, which are partially fused with the basisphenoid, underlie ventrally the anteroventral surfaces of the tuberculi basillare. Traces of sutures of a thin plate of the parasphenoid body and the ventral surface of the basisphenoid are distinct in places in Suchogorgon (Ivakhnenko, 2005b, p. 412); however, in the majority of taxa under study, this membrane bone is either reduced or completely lack sutures. The plate of the rostrum parasphenoidale, which spreads under the rostrum basisphenoidale, is usually independent. In the majority of taxa, it is a thin sulcate plate; however, in some taxa (e.g., Alrausuchus: Fig. 28a; Biarmosuchus, Suchogorgon: Ivakhnenko, 2005, text-fig. 12), it becomes a massive, high vertical plate that enters the interpterygoid fissure. The lower part of the plate projects far ventrally, and its lateral surfaces have rugose crests, which resemble attachment areas for muscles. True connection between the palatal segment and the braincase is performed in Dinomorpha by the epipterygoids. In all taxa that have preserved this bone, the epipterygoid is in the shape of a very thin vertical band. This structure of the epipterygoid is characteristic of all Dinomorpha (see Ivakhnenko, 2005b, text-figs. 16, 23). The upper margin of the bone is slightly thickened and adjoins dorsally the facies epipterygoidei parietalis (sutura sphenoparietalis). In Suchogorgon, the plate of the epipterygoid has a small expansion (fovea obturatoria: Ivakhnenko, 2005b, text-fig. 12) below the sutura sphenoparietalis; structural features of this formation suggest that it came in contact with a membrane, which was probably similar to the membrana spheno-obturatoria of monotremes (Ivakhnenko, 2005b, p. 438). In other taxa, similar structural features have not been recorded. In the lower part, the narrow plate of the epipterygoid sharply expands, forming a massive basis (basis epipterygoideus), which entered the sulcus epipterygoideus of the pterygoid. The medial surface of the basis epipterygoideum (pars basicranialis) has a contact facet for the basipterygoid process of the basisphenoid; however, in the majority of taxa, structural features of this region remain uncertain. For available data on Gorgonopidae, see Ivakhnenko (2005b, p. 420); almost the PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

901

same structure of contact regions are observed in Syodon and Titanophoneus. In Delectosaurus, as is possible to judge, only the massive basis epipterygoideus well ossified in the region of the basipterygoid articulation. In Suminia, a part of the medial surface of the margin of the pterygoid just posterior to the incisura interpterygoidea probably also participates in the formation of the basipterygoid articulation. In Delectosaurus, the pterygoid margin medial to the epipterygoid base is sharply thickened, elevated, and overlies the anterolateral margin of the basisphenoid, forming a complex suture. The pterygoid probably also participates in the articulation in other taxa; however, tight contact and even frequent fusion of bones in this region complicate the resolution of this question. The caudal margin of the epipterygoid base varies in the extent of ossification in different taxa. The sulcus epipterygoideus of the pterygoid usually continues posteriorly on the processus quadrati and sometimes even expands caudally, embracing a massive cartilaginous continuation of the epipterygoid base, which leant on the anteromedial margin of the quadrate (Syodon, Titanophoneus, Deuterosaurus, Ulemosaurus, Suminia, Delectosaurus, etc.). In Alrausuchus, the caudal part of the epipterygoid base well ossifies and reaches the quadrate. In Gorgonopidae, Biarmosuchus, and Inostrancevia, the posterior margin of the epipterygoid base narrows and do not reach the quadrate (or reaches it by its narrow cartilaginous band or a ligament). The palatal and temporal segments of all Dinomorpha are articulated with the lower jaw by the quadrate– quadratojugal complex (QQJ-complex) (Ivakhnenko, 2005b, p. 418). The complex is formed of the quadrate and quadratojugal. The tightness of connection between these bones varies in different taxa, and possibly depends on individual age of animals. The connection between the bones is in the shape of a sulcate apophysis, because a membrane bone comes in contact with an endochondral bone. However, particularly in large forms, the suture between bones is often obliterated. The paraquadrate canal passes above the mandibular condyle between the quadrate and quadratojugal and opens posteriorly in an ovate foramen (foramen quadratum). A similar foramen is at the anterior end of the canal that opens in a flat depression (facies anterior quadrati). The function of this canal remains uncertain; it was proposed that it may be connected with the sound-conducting system (Ivakhnenko, 2005b, p. 437). It is interesting that, in Deuterosaurus (specimen PIN, no. 1954/1), the canal seems closed; in my opinion, this makes groundless the hypothesis that this canal provided passage for a large blood vessel or a nerve. In primitive taxa (e.g., Reiszia: Fig. 17a) and taxa retaining streptostyly, (Biarmosuchus, Gorgonopidae, Inostrancevia, Viatkogorgon), the upper margin of the quadrate is thickened, forming a more or less massive capitulum quadrati with a parachondral surface. This capitulum quadrati enters the fossa quadratica squa-

902

IVAKHNENKO

(b) cqu

cic

com

lt

col

(c)

ciar

(a) car pep cria pra

Fig. 18. Quadrate–articular region of Reiszia gubini Ivachnenko, 2000, holotype PIN, no. 162/32: (a) right quadrate, medial view; (b) right quadrate, posterior view, articular region; and (c) right articulare, dorsal view. Designations: (car) cavum articulare of the articular bone, (ciar) cavum infraarticularis of the articular bone, (cic) crista intercondylaris of the quadrate, (col) lateral condyle of the quadrate, (com) medial condyle of the quadrate, (cqu) capitulum of the quadrate, (cria) crista interarticularis of the articular bone, (lt) lamina tenticulata of the quadrate, (pep) epipterygoid process of the quadrate, and (pra) retroarticular process of the articular bone. Scale bar, 0.5 cm.

mosa, providing mobility of the QQJ-complex (Ivakhnenko, 2005b, p. 437). As indicated above, in taxa that have lost streptostyly, the fossa quadratica is flattened and the pterygoid and (or) epipterygoid lean on the anterior margin of the QQJ-complex. The quadratojugal is usually flattened, tightly adjoining the anterior surface of the fossa quadratica (Fig. 21d) and often gets a complex, wavy contact surface. This is particularly well pronounced in Dicynodontia, in which the quadratojugal expands so greatly that exceeds in height the quadrate. However, the absence in this region of a true sutural articulation often results in the isolation of the QQJ-complex even in representatives of Ulemicia, Dicynodontia. The medial margin of the plate of the quadrate forms a flank (lamina tenticulata), with the posterior surface positioned obliquely relative to the sagittal plane. In Sauroctonus and Suchogorgon (Ivakhnenko, 2005b,

text-fig. 30a), the posterior surface of the lamina is almost flat, even, with a narrow groove (sulcus epipterygoideus) for the cartilaginous plate of the epipterygoid in the lower part. Reiszia (Fig. 18a, pep), Biarmosuchus, and Alrausuchus (Fig. 39b) are similar in the structure of these bones; in fact, below the sulcus epipterygoideus, they lack a special even contact area for a flat quadrate ramus. In Estemmenosuchus (Fig. 15a), Archaeosyodon (Fig. 15b), Syodon, Ulemosaurus (Fig. 14a), and Titanophoneus (Fig. 39c), the lower part of the surface of the flank is widely concave (forming sulcus stapedialis); below it, there is a flat contact facet for the pterygoid. In all Anomodontia examined (Figs. 13b, 14b, 39d), the posterior surface of the flank is flat; only the lower part, below the level of the squamosal, has a narrow groove (sulcus stapedialis) and a contact area for the pterygoid under it. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

903

(d) Ar

San (a) com

Qu (c)

com col

San

car

(b)

com cic col ciar

Fig. 19. Quadrate–articular region of the Biarmosuchus tener Tchudinov, 1960, specimen PIN, no. 1758/86: (a, b) right quadrate, articular region: (a) posterior and (b) anteroventral views; (c) right articular bone, dorsal view; and (d) quadrate and articular bone in articulation, dorsal view. Designations: (Ar) articular, (car) cavum articulare of the articular bone, (ciar) cavum infraarticularis of the articular bone, (cic) crista intercondylaris of the quadrate, (col) lateral condyle of the quadrate, (com) medial condyle of the quadrate, (Qu) quadrate, and (San) surangular. Scale bar, 1 cm.

The block of the quadrate–articular articulation is always divided into the lateral and medial condyles. In the Nikkasauridae (Fig. 18b), the articular surface of the quadrate is narrow and elongated, the surface of the lateral condyle is convex, cylindrical. It is separated by a weak flat groove (eminentia intercondylaris) from the somewhat elevated medial condyle, the lower surface of which is widened, considerably flattened, and curved posteriorly. In Alrausuchus, the lateral condyle is enlarged and retains a cylindrical shape. The eminentia intercondylaris becomes a high vertical plate and, hence, the flattened medial condyle is highly elevated and reaches the articular bone only when the lower jaw occupies the extreme posterior position. In this case, PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

the role of the working condyle is completely transferred to the cylinder of the lateral condyle. Biarmosuchus (Figs. 19a, 19b) shows the same structural pattern of the condyles; however, the eminentia intercondylaris is relatively much lower, and the lower surface of wide medial condyle is flat. The articular surface of the quadrate of Gorgonopidae (Ivakhnenko, 2005b, text-figs. 30a, 30b) shows the same structure. However, the statement that Suchogorgon lacks division into the lateral and medial condyles (Ivakhnenko, 2005b, p. 418) is in error. Comparisons with the taxa described show that the element designated the tuberculum articularis in fact corresponds to the medial condyle. The difference from the articulation of Biarmosuchus is restricted to the fact

904

IVAKHNENKO (b) (a)

cic

car ciar

com

cria

col

(c)

ciar

pra San (d)

pra

tms

Fig. 20. Quadrate and elements of the lower jaw: (a) Syodon efremovi (Orlov, 1940), specimen PIN, no. 157/670, right quadrate, anteriorly, articular surface; (b) Titanophoneus potens Efremov, 1938, specimen PIN, no. 157/221, right articular bone, articular surface; (c) Inostrancevia sp., specimen PIN, no. 2005/2265, left articular bone, medial view; and (d) Syodon efremovi (Orlov, 1940), specimen PIN, no. 157/677, symphyseal region of the lower jaw, lingual view. Designations: (car) cavum articulare of the articular bone, (ciar) cavum infraarticularis of the articular bone, (cic) crista intercondylaris of the quadrate, (col) lateral condyle of the quadrate, (com) medial condyle of the quadrate, (cria) crista interarticularis of the articular bone, (pra) retroarticular process of the articular bone, (San) surangular, and (tms) mental tubercle of the splenial. Scale bar, 1 cm.

that the eminentia intercondylaris is somewhat lower and the medial condyle is smaller in relative area. In other Dinocephalia (Archaeosyodon, Syodon, Titanophoneus, Ulemosaurus, Deuterosaurus), the eminentia intercondylaris looks like a superficial groove, the medial condyle is lowered somewhat more than the lateral condyle, both are in the shape of convex cylindrical surfaces positioned obliquely relative to the plane of the quadrate (Fig. 20a). The longitudinal axes of the condyles correspond to the articulation line of the lower jaw. The same design of the articular surface of the quadrate is in Estemmenosuchus. This region of Ulemicida has approximately the same structure. In Ulemica (Fig. 21a), the condyles are more rounded isometric and less convex than in typical Dinocephalia, and the lower side of the lateral margin of the quadratojugal (apophysis supracondylaris) is flattened, adjoins respective area on the

surangular (Figs. 21b, 21c); this interferes with lateral shifts of the jaw. In Suminia, the eminentia intercondylaris is cut highly and, hence, both condyles are only poorly convex lateral walls of the eminentia intercondylaris. This design reaches extreme development in the higher Dicynodontia (for example, in Vivaxosaurus, Fig. 21d). 5. Lower Jaw (ossa mandibularia) In the typical case, each lower jaw ramus of Dinomorpha includes the articulare, supraangulare, angulare, praearticulare, coronoideum, spleniale, and dentale. The articulare is ossification in Meckel’s cartilage, which adjoins neighboring membrane bones according to the apophysis type, membrane bones are usually articulated with each other by flat sutures (sutura plana). In the PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA) QQJ

(c)

905

(d)

(a)

aps cic

com

col

ta ciar com

col ciar

car

cria

pra (b) pra pra

(e)

car

Fig. 21. Quadrate and elements of the lower jaw: (a–c) Ulemica invisa (Efremov, 1938): (a) right quadrate, specimen PIN, no. 157/668, posterior view; (b) right articular bone, specimen PIN, no. 157/1112, dorsal view; and (c) the same elements in articulation, posterior view; (d) Vivaxosaurus permirus Kalandadze et Kurkin, 2000, holotype PIN, no. 1536/1, left quadrate, posterior view; and (e) Idelesaurus tataricus Kurkin, 2006, specimen PIN, no. 156/122, left articular bone, dorsal view. Designations: (aps) apophysis supracondylaris of the quadratojugal, (car) cavum articulare of the articular bone, (ciar) cavum infraarticularis of the articular bone, (cic) crista intercondylaris of the quadrate, (col) lateral condyle of the quadrate, (com) medial condyle of the quadrate, (cria) crista interarticularis of the articular bone, (pra) retroarticular process of the articular bone, (QQJ) quadrate–quadratojugal complex, and (ta) articular tubercle of the quadrate. Scale bar, 1 cm.

majority of taxa, the lower jaw functions as a united structure. In the Gorgonopidae (and Inostranceviidae), the jaw rami are divided into weakly connected dentary and postdentary parts (Ivakhnenko, 2005b). The major functional regions of the lower jaw of Dinomorpha are the articulation region (composed of the ossa articularia and, partially, ossa supraangularia), auditory region (ossa angularia = ectotympanicum, partially, probably, ossa praearticularia = gonialia), and the region of food treatment and insertion of jaw muscles (ossa dentalia and, partially, ossa supraangularia). The other bones are used for connection of the main zones. The articular surface of the articulare of all Dinomorpha is divided into two parts, the lateral cavum articularis and the medial cavum infraarticularis, which are separated by a more or less developed crista interarticularis. In the Nikkasauridae (Fig. 18c), the crest is very weak, but the cavum articularis is narrower and cylindrical, while the cavum infraarticularis is more flattened and expanded. In Alrausuchus, the cavum articularis has a concave cylindrical surface, and the crista interarticularis is in the shape of a ledge, with a flattened surface of the cavum infraarticularis lowered PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

such that it does not reach the surface of the medial condyle of the quadrate. The articular surface of the bone of Biarmosuchus (Fig. 19c) is similar in structure; however, its ledge forms the crista interarticularis relatively much lower; therefore, a wide flat surface of the cavum infraarticularis tightly adjoins from below a similar flat surface of a wide medial condyle of the quadrate; this is evidence that mobility in this articulation is impossible (Fig. 19d). In addition, the anterior upper margin of the cavum articularis forms the posterior margin of the surangular. In Alrausuchus, this bone margin forms a wide groove, which contains the lateral part of the articular bone. In Biarmosuchus, the upper margin of the groove curves somewhat posteriorly (tuberculum anterior ossis supraangularis), coming to the anterior surface of the lateral condyle and preventing movements of the quadrate. The same design of the articulation of the articular bone is characteristic of the Gorgonopidae (Ivakhnenko, 2005b, text-figs. 30c, 30d), in which the surface of the cavum infraarticularis is also flat and tightly adjoins the surface of the medial condyle (fossa quadrata: Ivakhnenko, 2005b). At the same time, the tuberculum anterior of the surangular is

906

IVAKHNENKO

developed much more strongly, forms the entire anterior part of the articular surface and, along with the projection of the posterior margin of the cavum articularis, sharply constricts the upper gap of the articular groove, preventing the movement of the quadrate. In Dinocephalia lacking streptostyly (Archaeosyodon, Syodon, Titanophoneus, Ulemosaurus, Deuterosaurus), the narrower and elongated cavum articularis and the wider and rounded cavum infraarticularis are deepened and become obliquely longitudinal grooves with a depressed surface (Fig. 20b). The same general design of the articular surface of the articulare is observed in Estemmenosuchus. The same is true of Ulemicida. In Ulemica (Fig. 21b), the cavum articularis and cavum infraarticularis are superficial ovate depressions separated by a gently sloping crista interarticularis. The anterior margin of the cavum articularis is almost completely formed by a very high tuberculum anterior supraangularis; this corresponds to the primarily high position of the posterior margin of the surangular. In addition, the tuberculum anterior embraces laterally the cavum articularis, forming a flat vertical area (fossa lateralis), which adjoins a flat external margin of the apophysis supracondylaris quadgatojugalis to prevent lateral movements of the jaw. At the same time, the posterior margin of the cavum articularis is rather flat, open posteriorly, and virtually lacks tuberculum posterior (Fig. 21c). In the Galeopidae (Suminia), the articular surface of the articular bone is changed considerably. The basic working structure of this bone is a very high crista interarticularis, which corresponds to the deep eminentia intercondylaris quadrati. The flattened cavum articularis and cavum infraarticularis are formed by lowering lateral slopes of this high rounded crest. This design of the articular surface is observed in all higher Dicynodontia (see, for example, Idelesaurus, Fig. 21e). The posterior margin of the articulare continues in the retroarticular process. In the Nikkasauridae (Fig. 18c), it is a small rounded tubercle with a depressed parachondral posteroventral surface and completely smooth upper surface. Almost the same structure of this process is observed in the majority of taxa (Alrausuchus, Syodon, Titanophoneus, Biarmosuchus, Ulemica, Suminia, etc.). In the Dicynodontidae, in connection with the expansion of the articular surface, the process almost lacks a dorsal surface. In Inostrancevia, the retroarticular process is particularly interesting; it is massive, sharply curves downwards, its rounded dorsal surface is smooth; laterally, the process is almost completely covered by the angular bone. Its ventral margin and a part of the posterior margin have distinct traces of cartilaginous continuation (Fig. 20c). In the Gorgonopidae, the retroarticular process is very thin and long, curves almost vertically downwards, and reaches the lower margin of the ala angularis; the lower margin is somewhat thickened and gently curves anteriorly at an almost right angle (Ivakhnenko, 2005b, text-fig. 30c). The posterior surface of the retroarticular

process is smooth, rounded, and the ventral surface is even, slightly rough, could have been covered by cartilage. The structure of the retroarticular process of Dinomorpha differs from a structure connected with attachment of depressor muscles, even if these muscles were poorly developed. A detailed analysis of this question has shown that the presence of a postquadrate depressor of the reptilian type in Theromorpha as a whole is highly improbable (Ivakhnenko, 2005b, p. 441). The structural features described above in this region of various Dinomorpha corroborate this conclusion, although no additional data on the true position of the depressor are provided. Medially and lateroventrally, the articular bone is covered by thin plates of the prearticular, surangular, and angular, respectively. These bones form the border of the mandibular fossa. In fact, the mandibular fossa as a structure containing the jaw muscles is not developed even in the Nikkasauridae (Fig. 22a). The upper margin of the prearticular tightly adjoins the internal surface of the lateral wall in the region of the suture between the surangular and angular and, hence, the muscles could only be attached to the upper part of the internal surface of the surangular. This surface is smooth and slightly concave. In some taxa (Reiszia, Syodon, Alrausuchus: Fig. 21b), the lower edge of the bone on the medial surface is bordered by a more or less developed rough crest, which delineates from above a groove (sulcus angularis superior). The groove expands somewhat posteriorly and comes onto the surface of the surangular, forming a weak incisure in the lower part of the posterior margin of the surangular. In this area, the groove is separated by a crest from the articular. The lower margin of the groove throughout its extent passes into a longitudinal depression (fovea praearticularis) on the medial surface of a thin plate of the prearticular, which tightly adjoins the angular. The anterior margin of the prearticular passes anteriorly, and its lower margin is separated by an incisure (incisura angularis medialis) from the upper margin of the angular bone. The shape of the incisure is extremely changeable; sometimes, it is almost absent (in Gorgonopidae), looks like a small foramen (for example, in Syodon: Fig. 23a), or becomes a large fenestra (in many Anomodontia, for example, in Suminia, Fig. 24b). The function of this structure is uncertain. The upper margin of the surangular curves morphologically downwards, lowering the region of the articular articulation. In the Nikkasauridae, this angle is the smallest (deviation from the longitudinal axis of the upper margin of the jaw is about 20°); in Phthinosaurus, Dinosaurus, Alrausuchus, and Dicynodontidae, it is approximately twice as great; in Gorgonopidae, Inostrancevia, Biarmosuchus, and Suminia, the angle is 50°; in taxa with a well-pronounced klinorhiny of the skull (Estemmenosuchidae, Tapinocephalida) and taxa with a very short and lowered postdentary part of the lower jaw (Syodon, Ulemica), the angle ranges from 65° to 80°. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

907

(a) Cr

Pcr

aan (b) sas

(c) sia

aan cia

Fig. 22. Left ramus of the lower jaw, lingual view: (a) Reiszia gubini Ivachnenko, 2000, reconstruction based on holotype PIN, no. 162/32; (b) Alrausuchus tagax (Ivachnenko, 1990), reconstruction based on holotype PIN, no. 3706/10, specimens PIN, nos. 3706/17, 3586/14, and 4659/8; (c) Biarmosuchus tener Tchudinov, 1960, reconstruction based on specimens PIN, nos. 1758/86 and 307. Designations: (aan) ala angularis, (cia) crista interlaminata of the angular bone, (Cr) coronoid, (Pcr) precoronoid, (sas) sulcus angularis, and (sia) sinus angularis. Scale bar, 1 cm.

The anterior margin of the mandibular fossa is formed by a thin plate of the coronoid. The most primitive taxa (Reiszia: Fig. 22a; Alrausuchus: Fig. 22b; Biarmosuchus: Fig. 22c) have an additional narrow second bone (praecoronoideum, or coronoideum anterior after Romer and Price, 1940). In Syodon (Fig. 23a), the coronoid is probably reduced, because it has not been recorded in four available lower jaw rami. As the precoronoid is present, its upper margin forms a crest bordering lingually the tooth area of the dentary. In other taxa, this function is performed by the upper margin of the increased splenial. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

The lower margin of the angular bone of some taxa is sharpened and forms a thin external crest (lamina reflexa, Fig. 23b, lr). The external crest of Gorgonopidae, Alrausuchus, Biarmosuchus, and Estemmenosuchus is well developed. In Nikkasaurus, Proburnetia, Syodon, Titanophoneus, Ulemosaurus, and Suminia, the crest is very short or absent. If the crest is developed, the anterior part of its internal surface is probably connected to the place of attachment of some portions of the pterygoid muscle (see Ivakhnenko, 2005b, p. 442). The most important structure of the angular is the ala angularis. In all taxa, the external surface of the

908

IVAKHNENKO

(a) sas

aan (b)

cia

lr

Fig. 23. Left ramus of the lower jaw, lingual view: (a) Syodon efremovi (Orlov, 1940), holotype PIN, no. 157/2; and (b) Estemmenosuchus uralensis Tchudinov, 1960, reconstruction based on specimens PIN, nos. 1758/22, 200, and 227. Designations: (aan) ala angularis, (cia) crista interlaminata of the angular, (lr) lamina reflexa of the angular, and (sas) sulcus angularis. Scale bar, 3 cm.

angular bone has an oblique groove extending from the dorsoposterior to anteroventral point, the anterior margin of which curves sharply posteriorly, forming a thin and wide plate (ala angularis). The lower margin of the plate descends below the edge of the angular for a very short distance (Syodon: Fig. 23a), approximately onethird of its height (for example, Alrausuchus: Fig. 22b; Suminia: Fig. 24b; and others), or even half of its height (Gorgonopidae; Estemmenosuchus: Fig. 23b). A flat fissura periangularis is formed between the bone body and ala angularis of the angular bone (Ivakhnenko, 2003b). The medial wall of this cavity is smooth, while the lateral wall (internal surface of the ala angularis) has an oblique crest (crista alata) (Fig. 25a, ca). This crest separates the upper part of the cavity (fissura externa, fex) from the lower part (fissura interna, fint). Upwards, the narrow fissura externa expands conically and forms a rounded incisure (recessus alae) in the anterodorsal margin of the wing; downwards, it narrows and passes into an extensive fissura interna. Below the ventral margin of the angular, the fissura interna passes into the cavity on the depression of the medial surface of the prearticular and is separated from the pterygoid fossa (if it is developed) by an oblique crest

(crista interlaminata) (Figs. 22c, 23b, cia). The margin of the recessus alae is always smooth, while the other margins of the wing have a more or less pronounced scalloped pattern, indicating connection with connective tissue. In a number of taxa (Estemmenosuchus, Syodon, Titanophoneus, Ulemosaurus), the ala angularis is massive and thick, its lateral surface is almost smooth and slightly convex. The other taxa investigated have a thin ala angularis; the crista alata becomes a more or less widened ridge, which corresponds to a relatively wide groove on the external surface, and the surface of the wing becomes wavy. This wing design is observed in Nikkasauridae, Alrausuchus, and Biarmosuchus; in Inostrancevia, it is particularly well pronounced (Fig. 25b). In the representatives of Anomodontia examined, this wing is somewhat thicker, and its surface is less wavy, except for Ulemica (Fig. 24a) and Suminia (Fig. 24b), with a thin, slightly wavy wing. In the higher Dicynodontia (for example, Idelesaurus and Dicynodon), the angular wing deviates gently laterally, and the upper cavity becomes a wide pocket. In these taxa, the lamina reflexa consists of only crista interlaminata. On the contrary, in Suchogorgon, Sauroctonus, PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

909

(a)

aan (b)

aan Fig. 24. Lower jaw, lingual view: (a) Ulemica invisa (Efremov, 1938), specimen PIN, no. 157/1112; and (b) Suminia getmanovi Ivachnenko, 1994, specimen PIN, no. 2212/62. Designations: (aan) ala angularis. Scale bar, 1 cm.

and probably all Gorgonopidae (Ivakhnenko, 2005b, p. 429), the same general structural pattern is combined with a very thin ala angularis, the upper part of which closely approaches the surface of the angular, and the recessus alae is very weak and flat. The structure of the fissura periangularis varies in different taxa only in relative area, extent of vertical or horizontal elongation, relative sizes of the upper and lower parts, extent of deepening, and the size of the recessus alae. These differences are connected with the shape of the cavity; however, the significance of these variations remains uncertain. In all cases, the incisure in the upper margin of the wing (recessus alae) opens dorsally and posteriorly, passing into a distinct gently sloping groove, which narrows sometimes posteriorly and is usually covered from below by the upper margin of the wing (Reiszia, Alrausuchus, etc.); sometimes, it is wide, open posteriorly (Anomodontia, Inostrancevia). An important role of this structure is supported by the fact that the region of the recessus alae is usually covered dorsoanteriorly by an overhanging margin of the lowered zygomatic arch, which is usually thickened externally in this area (sometimes pachyostotic) and smooth on the internal surface. Even in taxa without pachyostotic expansions, PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

this region of the lower margin of the arch is thickened (Nikkasaurus, Alrausuchus, Biarmosuchus, etc.). In Titanophoneus (Fig. 26) and Estemmenosuchus mirabilis, the zygomatic arch deviates somewhat farther from the lateral surface of the bone than in other taxa, and a small external portion of the adductor probably came onto the external surface of the posterodorsal margin of the dentary. The area for this muscle is separated from the region of the recessus alae and wing by a welldeveloped oblique rugose crest on the surface of the angular along the line of contact with the dentary (Fig. 26, see also Ivakhnenko, 2003c, text-figs. 13a, 16). In Anomodontia, the zygomatic arch of which curves upwards, the region of the recessus alae is usually covered from above by a greater or lesser developed longitudinal crest on the surangular and on the posterior margin of the dentary, which borders from below the area of the external muscle. Along the suture with the surangular, the medial surface of the upper margin of the angular almost always has a more or less developed groove (sulcus angularis superior) described above (Figs. 22b, 23a, sas). The groove is fused by its lower margin with a flattened surface of the prearticular and passes ventrally into the fissura interna. In Biarmosuchus, the posterior part of the

910

IVAKHNENKO (b)

ca

fex

ra

fint (a)

ra fex

ca fint

Fig. 25. Internal surface of the ala angularis: (a) Syodon efremovi (Orlov, 1940), holotype PIN, no. 157/2; and (b) Inostrancevia sp., specimen PIN, no. 2005/1707. Designations: (ca) crista alata of the angular, (fex) fissura externa of the angular, (fint) fissura interna of the angular, and (ra) recessus alae of the angular. Scale bar, 1 cm.

Sq

Ju

Mx

QQJ

Fig. 26. Quadrate-articular region of the skull of Titanophoneus potens Efremov, 1938, lectotype PIN, no. 157/1, right lateral view. Designations: (Ju) jugal, (Mx) maxillary, (QQJ) quadrate–quadratojugal complex, and (Sq) squamosal. Scale bar, 3 cm.

groove expands into a vertically elongated depression (sinus angularis, Fig. 22c, sia), which corresponds to a low vertical eminence (eminentia angularis) on the lat-

eral surface of the angular. In the Gorgonopidae, this eminence corresponds to a narrow crest (Ivakhnenko, 2005b, p. 429: spina angularis), which extends on the PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

external surface of the angular just anterior to the place of deviation of the ala angularis. In the upper part, the margins of the crest are almost vertical, becoming gently sloping downwards. The upper margin of the crest has one or two small foramina of canals extending inside the angular. Ivakhnenko (2005b, p. 442) proposed that the spina angularis and its foramina could have been related to certain structures of the external ear. The firmness of attachment of the dentary to postdentary bones varies in different taxa. The lower surface of the caudal part of the dentary (ramus dentalis) is convex and lies in a more or less expressed groove formed by the anterior processes of the surangular and angular (processus intramandibularis), which adjoin each other. The medial side of contact between the margins of the dentary and angular has a foramen (foramen angularis lateralis). In taxa with a well-pronounced sculpture on the lateral surface of the dentary (Suchogorgon, Syodon, Ulemica, etc.), a groove, which often branches, extends anteriorly from the foramen on the surface of the dentary. In the majority of taxa under study, the foramen is relatively small, circular (Nikkasauridae, Ulemosauridae, Estemmenosuchidae, etc.), or sometimes slitlike (for example, in Alrausuchus). In Titanophoneus, the foramen is separated dorsally from the external portion of the adductor by a low crest. In Anomodontia, particularly in Dicynodontida, this foramen is very large. It is probable that vessel, maybe similar to the vena facialis of mammals, passed in this area (see Ivakhnenko, 2005b, p. 444); it most likely fed only tissues of the dentary surface. This agrees with a changeable size of the foramen, which is increased in Dicynodontida because of longitudinal movements of the lower jaw. Below the foramen, many taxa have a small projection of the posterior margin of the dentary, which borders the foramen from below (processus angularis). Sometimes, this process is excessively developed, passes posteriorly on the surface of the angular (in Dicynodontida, and also in Ustia: Ivakhnenko, 2003c, text-fig. 21b), providing additional reinforcement of connection between the dentary and postdentary region. On the lingual surface, the two parts of the jaw are mostly connected by the splenial. This bone is connected to the angular by a lamellar suture, overlies lingually the sulcus mylohyoideus, and participates in the formation of the lower jaw symphysis. In the symphyseal region, the bone is most tightly connected to the dentary through areas with longitudinal crests. The splenial is often fused with its counterpart, and the symphyseal suture is sometimes completely obliterated (Syodon, Titanophoneus, Ulemosaurus, Ulemica, Dicynodontidae). The plate of the splenial usually narrows before the symphysis; as a result, the anterior margin of the sulcus mylohyoideus is exposed. In the Gorgonopidae (Ivakhnenko, 2005b, p. 430), the splenials are connected by the considerably thickened lingual processes (pars symphyseos), with sharp, PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

911

high, serrated bifurcating crests on the facies symphyseos. The attachment to the dentary is provided by a circular tubercle (tuberculum mentalis), which enters a fossa (impressio mentalis ossis dentalis). This provides rigid attachment of both postdentary regions of the jaw and mobile articulation between these regions and dentaries. Unfortunately, this region is insufficiently examined, and available data are only restricted to a few taxa. In Suchogorgon, the lingual side of the splenial posterior to the pars symphyseos has a fossa and a thickened tubercle (Ivakhnenko, 2005b, text-fig. 31a). Similar structures are recorded in almost all taxa investigated (except for Nikkasaurus, the symphysis of which is low and elongated). The structure often looks like symmetric fossae separated by a more or less developed crest (tuberculum mentalis) (Fig. 20d, tms). The presence of this structure, which is probably related to certain differentiation of premaxillary muscles, is probably attributable to the attachment in this region of analogues of the depressor muscles of mammals (Ivakhnenko, 2005b, p. 443). The dentaries adjoin each other at the facies symphyseos, forming the synchondrosis intermandibularis. On this surface, weak low radiating crests of contact extend from the region of the foramen mylohyoideus anterior. In Reiszia, Biarmosuchoides, and Ustia, the facies symphyseos are longitudinally extended, have a smooth surface; the articulation was probably relatively weak. In many taxa (Syodon, Titanophoneus, Ulemosaurus, Deuterosaurus, and Anomodontia examined), the dentaries are fused, often with a completely obliterated suture (at least, at the definitive stage). In the majority of taxa, the dorsal surface of the dentary has depressions (alveolae dentaries) arranged in a row extending for some extent from the symphysis. In the Nikkasauridae, the anterior part of the symphyseal region lacks teeth (Fig. 22a). In Venyukovia, Ulemica, Niaftasuchus, Suminia, and Microurania, from one to three anterior alveoli are increased (alveoli dentum incisivorum); subsequent alveoli (alveoli dentum buccinorum) gradually decrease in size posteriorly. In Deuterosaurus and Microurania, the incisor and cheek alveoli are separated by an increased alveolus of the canine-shaped tooth. In groups that have true lower canines, the fifth or sixth anterior alveolus is very strongly increased, forms alveolus for the canine (alveolus dentis canini inferioris). In the Gorgonopidae, Inostrancevia, Biarmosuchus, and Titanophoneus, the canine alveolus contains a narrow additional alveolus of the capsule of the replacement canine (bursa canini inferioris successorum) in the posterolateral corner. In the Gorgonopidae (Ivakhnenko, 2005b, text-fig. 32a), posterior to the canine capsule, there is a diastema with a row of four–six rudimentary alveoli, which corresponds in position to the upper canine. Posterior to the diastema (or posterior to the alveolus of the canine), there is a row of alveoli for postcanines (alveoli dentum

912

IVAKHNENKO (a) (b)

Fig. 27. Dorsal surface of the lower jaw: (a) Syodon efremovi (Orlov, 1940), holotype PIN, no. 157/2, last cheek teeth of the left ramus; and (b) Estemmenosuchus mirabilis Tchudinov, 1968, holotype PIN, no. 1758/6, symphyseal region. Scale bar, 1 cm.

postcaninorum). Inostrancevia lacks postcanines or traces of their alveoli. An unusual structure of the dorsal surface of the dentary is observed in Ulemica. The cheek teeth are arranged in a row extending along the lingual part of the surface, and the lateral part is rugose and considerably thickened; therefore, not only a fossa for the increased upper tooth, but also fossae of the apices of other upper cheek teeth are imprinted on the surface. In all taxa possessing cheek teeth (or postcanines), the dorsoexternal margin of the dentary is more or less elevated, forming the limbus dentalis. In the higher Dicynodontida, the anterior alveoli are absent, the expanded surface of the symphyseal region becomes porous, and the limbus dentalis forms a more or less expressed cutting border. In primitive taxa, the cheek alveoli are retained; for example, Australobarbarus has up to eight alveoli on an elongated tubercle; three or four anterior alveoli are large, while posterior alveoli are smaller and probably lack definitive teeth. Emydops has only two alveoli (Cluver and King, 1983, text-

fig. 33); the same is true of an undetermined dicynodont (specimen PIN, no. 4549/23). In higher dicynodonts, postcanine alveoli disappear, and the entire dentary surface has the same structure as the symphyseal region. In two cases, the alveoli of the lower cheek teeth are oblique. In taxa with a well-developed streptostyly (Gorgonopidae, Biarmosuchus), the inclination is weak, while, in taxa with well-developed klinorhiny of the skull (Ulemosaurus, Deuterosaurus, Rhopalodontidae, Estemmenosuchus), it is greater. In some taxa, the caudal cheek alveoli are occasionally positioned by two in a row (Syodon, Fig. 27a) or even by three (Ulemica). In Estemmenosuchus (Fig. 27b), the row of cheek alveoli curves lingually around the canine alveolus. The lingual surface of the dentary has a sulcus mylohyoideus, which always reaches the symphysis and gives rise to branches extending deep into the alveoli. Posterior to the alveolar line, the entire upper surface of the ramus dentalis has distinct traces of attachment of muscles on both lateral and lingual sides. In PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

Titanophoneus and Estemmenosuchus mirabilis, the attachment areas for muscles expand onto the lateral surfaces and reach the crests on the upper margin of the angular. In Anomodontia, the lateral surface of the ramus dentalis has a deep depression (“lateral shelf” after Cluver and King, 1983), which is bordered ventrally by a crest. This depression, along with a depression on the ventrolateral margin of the squamosal, above the QQJ-complex, form a structure resembling an attachment area of the detrahens mandibulae muscle of Monotremata. The Gorgonopidae have a well-developed apophysis coronoideus, which extends posteriorly farther than the contact area for the surangular (Ivakhnenko, 2005b, text-fig. 32a). In other taxa, the coronoid process is either in the shape of a small tubercle or absent. In many taxa, the lateral surface of the dentary is covered with the same sculpture as the facies facialis of the maxillary. Distinctions include the extent of development of grooves for blood vessels, which come from below out of the region between the jaws, curve around the lower margin of the dentary, branch and anastomose on the mental surface (facies mentalis) (Gorgonopidae, Biarmosuchus, Estemmenosuchus, Syodon, Titanophoneus, etc.). In primitive taxa (Biarmosuchus, Gorgonopidae, Ustia, Biarmosuchoides) the background sculpture is supplemented by circular pits. In Ulemica, a well-developed thickened pachyostotic crest (Fig. 24a) extends obliquely from the ventroexternal margin of the mental region to the midlength of the ventral surface of the bone (see also Efremov, 1940b, p. 55, textfigs. 16A, 16B). 6. Ethmoidal and Braincase Regions (endocranium) In Dinomorpha, many sites of this regions are represented by membrane formations or nonossifying cartilage. In connection with natural difficulties in the preparation of the nasal region, the ethmoid region is poorly understood. Only isolated data on a few taxa are available. The internasal septum (septum nasi osseum) probably ossifies to a greater or lesser extent in many taxa; it was recorded with certainty in the Gorgonopidae, Inostrancevia, Alrausuchus, and Biarmosuchus. In a number of taxa (Inostrancevia, Biarmosuchus: Fig. 16e), this septum has a relatively large foramen. In Ulemica, Suminia, Delectosaurus, and Dicynodon trautscholdi, ossification of the septum has not been recorded; possibly, it did not ossify in Anomodontia as a whole. In Suchogorgon, a special unpaired ossification designated as mesectethmoideum was described (Ivakhnenko, 2005b, p. 414); this is probably a partially ossified middle part of the planum antorbitale. The ethmoid region is relatively thoroughly examined in Suchogorgon (Ivakhnenko, 2005b, text-figs. 11– 14, 16, 17) and, judging from published data considered in this work, it is almost identical in all Gorgonopidae. It is composed of the unpaired bones mesethmoiPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

913

deum, orbitosphenoideum, praesphenoideum, and basisphenoideum and paired laterosphenoidea. The basisphenoid forms the anterior part of the bottom of the braincase. A thin plate of the presphenoid (ossification in the interorbital septum between the rostrum basisphenoidale and orbitosphenoideum) enters a groove on the dorsal surface of the rostrum basisphenoidale. The orbitosphenoid has lateral wings (alae orbitosphenoidei), which form lateral walls of the olfactory region of the braincase. This cavity bifurcates anteriorly, bypassing the mesethmoid plate. The posterior margins of the orbitosphenoid wings are thickened, have a parachondral surface and are cut deeply by the points of exit of the trochlear and oculomotor nerves. The cartilaginous plates of the lateral walls of the braincase, which continued the orbitosphenoid wings, adjoined the upper margins of the ossa laterosphenoidea and the anterior margins of the synoticum. The upper margin of a thin laterosphenoid bone comes in contact with the processus dorsalis of the prootic, forming an oval foramen for branches of the trigeminal nerve (Fig. 36a). Alrausuchus is the second taxon examined with reference to this region (Fig. 28a). The basic differences consist in the much lower olfactory region and, hence, relatively higher interorbital septum. The last character is connected with a high and thin plate of the ossified cartilage (praesphenoideum), which, in addition, extends far anteriorly and spreads under the mesethmoid. In addition, the roof of the olfactory region, which connects the alae orbitosphenoidei, is ossified. Thus, this region in general corresponds to the single ethmoid ossification in the sphenacomorph Dimetrodon (Romer and Price, 1940, text-fig. 11). Only very poor data are available on Proburnetia. The inner mold of the type specimen is broken at the sagittal line (Fig. 28b); however, judging from imprints, this animal shows the same general structural pattern of the region, only the orbitosphenoid is very low and short, and the high narrow basisphenoid rostrum rises abruptly upwards. Judging from the imprint, the dorsal wings of the palatine and pterygoid do not reach the ossified region of the interorbital septum. In Ulemosaurus, the basisphenoid rostrum also rises abruptly upwards (Tatarinov, 1965), and its laterosphenoid region ossifies to a much greater extent than in the taxa considered above. The region of the commissura orbito-parietalis even better ossifies in Deuterosaurus, in which the bony plate connects the laterosphenoideum, processus dorsalis of the prootic, and ala lateralis of the synotic (Efremov, 1954, text-fig. 10). However, in this case, the laterosphenoid is also a relatively narrow bone. At the same time, it is possible that, in different taxa (and even different age stages of the same species), the extent of ossification of the cartilaginous wall in the laterosphenoid region varied widely; for example, Sigogneau (1970a, text-fig. 19a) showed a very wide laterosphenoid plate in Gorgonops torvus.

914

IVAKHNENKO (a)

Osp

Mth

Prs (b) rp pb V

Mth

poc

Osp

Prs

Fig. 28. Parasagittal sections of the braincase: (a) Alrausuchus tagax (Ivachnenko, 1990), holotype PIN, no. 3706/10, left lateral view; and (b) Proburnetia vjatkensis Tatarinov, 1968, scheme based on holotype PIN, no. 2416/1, left lateral view. Designations: (Mth) mesethmoid, (Osp) orbitosphenoid, (pb) basipterygoid process of the basisphenoid, (poc) paroccipital process, (Prs) presphenoid, (rp) rostrum parasphenoidale, and (V) foramen nervi trigemini (rami hyoidei et palatini). Scale bar, 1 cm.

The basisphenoid, which spreads under the anterior part of the braincase, is very similar in general structural plan in different taxa, in which it was possible to examine this bone. The bone has a more or less elongated anterior rostrum basisphenoidale, which underlies the region of the interorbital septum. On the margins of the base of the rostrum, there is the region of the basipterygoid articulation, which provides connection with the palatal segment. The bone is connected to the basioccipital by the synchondrosis sphenooccipitalis, which is frequently involved in a synostosis, although traces of a suture are usually distinct. On the posterior margin of the ventral surface, it usually forms well-pronounced sphenooccipital tubercles (tuberculum basil-

lare), the areas of attachment of a part of subvertebral cervical muscles. The posterior surfaces of tubercles have areas (impressio prooticalis) for articulation with the prootic. The dorsal (pituitary) surface of the basisphenoid is occupied by elongated hypophyseal fossa (fossa hypophyseos = sella turcica), the anterior part of which has foramina for the internal carotid artery, and the posterior margin is limited by the dorsum sellae turcici varying in height. These structures vary somewhat in different taxa, the major differences concern the basipterygoid articulation. In the Nikkasauridae, the basipterygoid processes (Figs. 29a, 29b, 30a, pb) are massive and elongated, with a wide rounded anterior parachondral surface. In Suchogorgon PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

915

fptt ps Ex

poc

fptt

cs stc

fo

fc

pb Psh

pb

(b)

(a)

Fig. 29. Braincase of Nikkasaurus tatarinovi Ivachnenko, 2000, specimen PIN, no. 162/31: (a) ventral and (b) dorsal views. Designations: (cs) crista sellaris, (Ex) exoccipital, (fc) fossa carotica, (fo) foramen ovale, (fptt) fenestra posttemporalis, (pb) basipterygoid process of the basisphenoid, (poc) paroccipital process, (ps) supraoccipital process of the synotics, (Psh) parasphenoid, and (stc) sella turcica. Scale bar, 0.1 cm.

and Sauroctonus, the processes are relatively very small, short, with narrow triangular articular areas (Ivakhnenko, 2005b, text-figs. 12, 14). In all Dinocephalia examined, this region shows the same general structure. The basipterygoid processes are short, flattened, subtriangular in cross section, articulated with the epipterygoid by narrow lateral margins and underlain by the pterygoid (Figs. 12a, 28a, 31a, 31b, 32d). In Syodon, this region shows a distinct structure (Figs. 35a, 35b), i.e., its flat processes are very slightly underlain by the pterygoid. In Ulemica and Suminia (Fig. 13a), the proPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

cesses are relatively long, and almost the entire lateral surface comes in contact with the margin of the pterygoid. In Dicynodon and Delectosaurus, the basipterygoid process is short, dorsoventrally flattened, with a somewhat expanded contact lateral surface, the lower part of which forms an apophysis with the pterygoid, and the upper part is almost completely fused with a wide base of the epipterygoid. The second important point of structural differences in this region between different groups is the angle at which the basisphenoid and basioccipital are articu-

916

IVAKHNENKO

cfl (a) oaa ccl cc (b)

ps

pd

IX–XI XII oap

V

fptt V

poc

rp

fo

pb

Fig. 30. Braincase of Nikkasaurus tatarinovi Ivachnenko, 2000, specimen PIN, no. 162/31: (a) right lateral view and (b) sagittal section. Designations: (cc) crus commune, (ccl) foramen crus canalis semicircularis lateralis, (cfl) canalis floccularis, (fo) foramen ovale, (fptt) fenestra posttemporalis, (oaa) orificium ampullae canalis semicircularis anterior, (oap) orificium ampullae canalis semicircularis posterior, (pb) basipterygoid process of the basisphenoid, (pd) dorsal process of the prootic, (poc) paroccipital process, (ps) supraoccipital process of the synotic, (rp) rostrum parasphenoidale, (V) foramen nervi trigemini (rami hyoidei et palatini), (IX–XI) foramen nervorum glossopharyngei, vagi et accessorii, and (XII) foramen nervi hyppoglossi. Scale bar, 0.1 cm.

lated. The pituitary surface (planum pituitaria) of the basisphenoid and the dorsal plane of the basioccipital (clivus), separated by the dorsum sellae turcici, are positioned in almost the same plane in many groups (see, for example, Figs. 30b, 36a, 37a, plate braincase). However, in Dinocephalia, the basisphenoid is positioned at an angle to the plane of the clivus. In Alrausuchus, the inclination is about 30° (Fig. 34a); in Biarmosuchus and Syodon, it is up to 60° (Figs. 33b, 34b); and in Ulemosaurus (and probably Tapinocephalida as a whole), it reaches 80° (klinorhinal braincase). The basicranial region of the neurocranium is composed of the endochondral bones synoticum, opisthotica, prootica, basioccipitale, supraoccipitale, and exoccipitalia, which are usually connected in a single structure (perioticum) by synchondroses, which are often obliterated. The periotic forms a massive single occipital ring, which is overlapped dorsally and posteriorly by the postparietals and tabulars. The occipital ring borders the foramen magnum, provides connection with

the cervical vertebrae, forms the posterior and partially posterolateral walls of the braincase, and forms the otic connection with the palatoquadrate region (with the QQJ-complex). The posterior surface above the foramen magnum provides attachment area for supravertebral cervical muscles; the anterior surface (on the lateral margins) is attachment for the adductors. A very important point in the structure of the periotic is the fact that it is surrounded by the otic bones (synoticum, opisthotica, and prootica) of the labyrinth of the inner ear. In all the taxa investigated, the periotic ring shows the same general design (for detailed description, using the example of Suchogorgon, see Ivakhnenko, 2005b, pp. 405–409). Variations in other taxa usually concern the extent of obliteration of synchondroses (which is extremely weak in places with the preservation of nonossifying zones in Nikkasauridae, while Dinocephalia and Dicynodontia show extremely strong fusion of bones), the general proportions, and, particularly, the extent to which the paroccipital process is developed. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

917

(a) Qu

(b)

Psh

S

pb Pt

Fig. 31. Braincase of Alrausuchus tagax (Ivachnenko, 1990), holotype PIN, no. 3706/10: (a) ventral and (b) dorsal views. Designations: (pb) basipterygoid process of the basisphenoid, (Psh) parasphenoid, (Pt) pterygoid, (Qu) quadrate, and (S) stapes. Scale bar, 1 cm.

The number and arrangement of foramina in the region of the fissura metotica (canalis jugularis, IX–XI nerves) between basioccipitale and opisthoticum vary widely; this structure may vary even on different sides of the same skull. The anterior surface of the periotic lateral to the braincase is formed of the anteriorly directed anterior lamina, which is formed mostly of the prootic. The lateral surface of the plate forming the planum temporale usually has distinct traces of the origin of muscles; in PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

large forms, these are rugose crests. In small forms (Nikkasauridae, Fig. 30a), the surface is almost smooth and shows a relatively large foramen, from which a groove extends to the region where the internal carotid artery passes onto the basisphenoid. This foramen probably provided passage for the palatomandibular ramus of the trigeminal nerve (V). A very rough bone surface in this area and the presence of several emissaria, which accompany the branches of the cerebralis media vein prevented the recognition of this foramen in Suchogorgon (Ivakhnenko, 2005b). The upper margin

918

IVAKHNENKO (c)

(b)

tst

cs (a)

ib

ib

Psh tb

fci fc

fap

(d) fph

pb

Fig. 32. Braincase of Biarmosuchus tener Tchudinov, 1960: (a–c) specimen PIN, no. 1758/85: (a) dorsal view; (b) region of the paroccipital process, ventral view; and (c) base of the left stapes; and (d) specimen PIN, no. 1758/19, parabasisphenoid region, anterior and somewhat ventral view. Designations: (cs) crista sellaris, (fap) foramen canalis rami palatini arteriae carotica, (fc) fossa carotica, (fci) foramen canalis caroticus, (fph) facies rostri parasphenoidei of the basispheniod, (ib) impressio basicapsularis, (pb) basipterygoid process of the basisphenoid, (Psh) parasphenoid, (tb) tuberculum basillare, and (tst) tuberculum prooticalis ossis stapedis. Scale bar, 1 cm.

of the anterior lamina at the point of contact with the synotic forms an incisure for the facial nerve (VII), which is bordered anteriorly by the dorsal process. This process is short and massive in Nikkasauridae; narrow and high, reaching the laterosphenoid in Gorgonopidae (Fig. 36a), Ulemosaurus, and Deuterosaurus; very poorly developed and probably nonossified in Biarmosuchus (Fig. 33a); and well developed in Alrausuchus (Fig. 34a). In Suminia (Fig. 37a), the process is short, but well developed; in Delectosaurus, it is in the shape of a short tubercle. In all higher Dicynodontia examined (Dicynodon trautscholdi, Delectosaurus arefjevi, and Australobarbarus kotelnichi), the upper margin of the anterior lamina is thin and sharp; possibly, the lateral walls of the braincase were completely replaced by a membrane. Anterior to the process, the anterior lamina has an articular area for the laterosphenoid (facies laterosphenoidalis: Figs. 33b, 37a, fal). A large oval

foramen, which probably provided passage for branches of the trigeminal nerve (V), is located between the dorsal process and laterosphenoid. Farther, the anterior margin of the anterior lamina adjoins the upper part of the lateral surface of the dorsum sellae turcici, forming along with it a short canal for the abducent nerve. In all the taxa investigated, the ossifying tectus posterius (os supraoccipitale) is fused with the ossifying tectus synoticus (os synoticum) in a single supraoccipitale, although borders between them are sometimes distinct, particularly in longitudinal breaks (Fig. 36a). The upper surface of the synotic part is usually underossified, retaining traces of an ascending cartilaginous process that enters a fossa formed by both parietals (recessus tecti synotici, rts). The lateral margins of the bone form flattened conical processes directed laterally. These processes attach the braincase to adjacent PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

ps

(a)

(b) fptt

919

pd cfl fal

poc

Ex

fo

pb

Fig. 33. Braincase of Biarmosuchus tener Tchudinov, 1960, specimen PIN, no. 1758/85: (a) left lateral view and (b) sagittal section. Designations: (cfl) floccular canal, (Ex) exoccipital, (fal) facies laterosphenoidalis of the prootic, (fo) foramen ovale, (fptt) fenestra posttemporalis, (pd) dorsal process of the prootic, (poc) paroccipital process, (ps) supraoccipital process of the synotic. Scale bar, 2 cm.

membrane bones. Each process (processus mastoideus synoticus = processus supraoccipitalis, ps) is covered posteriorly by a plate of the tabular bone; anteriorly, it is overlapped by a thin plate of the squamosal (lower part of the ala perioticalis, pars synoticus). The lower margin of the process covers the canalis posttemporalis. In the majority of the taxa investigated, a depression of the fovea floccularis enters the supraoccipital process on the internal surface of the braincase. In Inostrancevia, this fovea is completely absent (Fig. 36b). In the Dicynodontidae (Delectosaurus: Fig. 37b), it is in the shape of a very superficial depression, but in Suminia (Fig. 37a), it is relatively deep, extending somewhat downwards. In Nikkasaurus (Fig. 30b), the fovea is superficial, but very wide, occupies almost entirely the internal surface of the supratemporal, expanding onto the opisthotic. In Biarmosuchus, the floccular fovea is relatively small in area (Fig. 33b); in Alrausuchus, it is relatively much wider, but superficial (Fig. 34a). In Suchogorgon (Ivakhnenko, 2005b, textfig. 9) and Sauroctonus (Fig. 36a), the canal of the floccular depression is large, in the shape of a curving cone, extends almost vertically downwards and somewhat posteriorly; the synotic forms its lower and posteromedial walls. The greatest relative size of the floccular PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

cavity is observed in Syodon (Fig. 34b); the cavity extends downwards and posteriorly, as in the Gorgonopidae; however, its bottom is abruptly and irregularly widened and the lower part of the cavity closely approaches the posterolateral surface of the periotic rather than has a narrow conical bottom (Fig. 34c). As the cavity is well developed, its walls often have rough lamellar crests. The lateral projection of the paroccipital process formed mostly by the opisthotic is developed below the supraoccipital process and separated from it by the posttemporal fenestra. This process varies in relative length in different taxa, depending on the ratio between the widths of the periotic and skull. For example, in the Inostranceviidae, with the expanded parietal and temporal regions of the skull, the paroccipital process is very narrow and long, while in Ulemosaurus (Fig. 14a) and Deuterosaurus, with narrow and high skulls, this process is very short. The total width of the periotic probably depends on the proportions of the braincase and ear capsules and is mostly independent of the skull width, which substantially correlates with the design of the temporal region. The process usually expands distally and has a perichondral lateral surface, the lower part of which adjoins the quadrate, or, in taxa with a

920

IVAKHNENKO

P (b)

cant

(c) cfl cpas clat

pd cc clat

pd

cfl

IX–XI cfl XII

Lat

cc (a)

vs

Fig. 34. Sections of periotic structures: (a) Alrausuchus tagax (Ivachnenko, 1990), holotype PIN, no. 3706/10, right part, parasagittal section; (b) Syodon gusevi (Tchudinov, 1968), holotype PIN, no. 2505/1, right parietal and periotic, sagittal section; and (c) Syodon efremovi (Orlov, 1940), specimen PIN, no. 157/1047, horizontal section of the region of the floccular fovea, dorsal view. Designations: (cant) canalis semicircularis anterior, (cc) crus commune, (cfl) canalis floccularis, (clat) canalis semicircularis lateralis, (cpas) canalis semicircularis posterior, (Lat) laterosphenoid, (P) parietal, (pd) dorsal process of the prootic, (vs) vestibulum, (IX–XI) foramen nervorum glossopharyngei, vagi et accessorii, and (XII) foramen nervi hyppoglossi. Scale bars: (a) 0.5 cm and (b, c) 1 cm.

pronounced streptostyly, forms the medial wall of the fossa articularis quadratica (see, for example, Fig. 12a). The caudal surface of the distal margin is usually rounded and projects only slightly posteriorly, although it sometimes passes laterally, overhanging the fossa elliptica squamosi (see Estemmenosuchus, Fig. 15a; Delectosaurus, Fig. 14b). This projection is most developed and even forms a flat lateral flank (lamina postoccipitalis) in a number of Dinocephalia, such as Syodon,

Archaeosyodon (Fig. 15b), Deuterosaurus, and Ulemosaurus (Fig. 14a). The flank covers posteriorly a part of the fossa elliptica squamosi and overhangs the sulcus epipterygoideus of the quadrate. Below the floccular fovea, the medial surface of the periotic forms the lateral walls of the recessus labyrinthicus. The sutures between particular ear bones composing the periotic are usually almost indiscernible; they are vague in places even in Nikkasaurus, with a PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA) (a)

921

(b)

fptt

ps

fo

poc

pb

Fig. 35. Braincase Syodon gusevi (Tchudinov, 1968), holotype PIN, no. 2505/1: (a) dorsal and (b) ventral views. Designations: (fo) foramen ovale, (fptt) fenestra posttemporalis, (pb) basipterygoid process of the basisphenoid, (poc) paroccipital process, and (ps) supraoccipital process of the synotic. Scale bar, 1 cm.

very weak synchondrosis; therefore, it is better to consider the structure of the bony labyrinth as a single element. The medial wall above the level of the clivus does not ossify in any taxon under study. As a result, a subrectangular fenestra exposing almost completely the pars utricularis vestibuli is formed somewhat posterior to the dorsum sellae (Figs. 30b, 34b, 36a, 36b, 37a, 37b). Its anterodorsal corner has a depression (ampulla ossea) with a foramen of the canalis semicircularis anterior, which curves inside the bone around the fovea floccularis (Fig. 38a). Posterior to the floccular fovea, the canal is fused with the canalis semicircularis posterior and forms a short canal crus osseum commune, which opens in a foramen in the posterodorsal corner of the fossa utricularis vestibuli. Under this foramen, there is an elongated depression crus ossea ampullaria and the foramen canalis semicircularis posterior; somewhat below, there is a foramen of the crus osseum simplex canalis semicircularis lateralis. This canal opens in a depression of the ampulla ossea, which is located just below the orificium ampullae canalis semicircularis anterior. Below the region of the canales semicirculares ossei, the vestibulum expands somewhat, forming a PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

depression, which is divided into the upper oval (recessus ellipticus) and lower circular (recessus sphericus) parts separated from each other by a very weak crista vestibuli; this probably corresponds to the division of the membrane labyrinth into the utriculus and sacculus (Fig. 38b). Delectosaurus lacks a distinct crest in this region (Fig. 38c). The anterior wall of this region lacks distinct depressions that could have been associated with the recessus lagenaris. Below the utriculosaccular region, the vestibulum either opens just laterally (Gorgonopidae: Fig. 36a) in the fenestra vestibuli (regarded here as the external border of the porus acusticus, which is frequently wider) or extends somewhat ventrally and laterally (Suminia; Dinocephalia: Fig. 38b). As the tuba vestibuli surrounding the oval fenestra is developed (Dicynodontidae: Fig. 38c), the lower part of the vestibulum sharply narrows, forming a narrow canal, and then opens in a small foramen (porus acusticus) in a wide conical depression of the oval foramen. The ventral wall of the oval foramen is often nonossified (Figs. 31a, 31b, 35a). The oval fenestra is usually surrounded by a depression (impressio basicapsularis: Figs. 32a, 32b, ib), which contained the base of the

922

IVAKHNENKO

ldl

(a)

pd

(b)

Syn cfl

cc Lat VI

Ex XII

cs

Fig. 36. Periotics in sagittal section, right half: (a) Sauroctonus progressus (Hartmann-Weinberg, 1938), holotype PIN, no. 156/51; and (b) Inostrancevia uralensis Tatarinov, 1974, reconstruction based on holotype PIN, no. 2896/1 and specimen PIN, no. 2896/3. Designations: (cc) crus commune, (cfl) canalis floccularis, (cs) crista sellaris, (Ex) exoccipital, (Lat) laterosphenoid, (ldl) lamina dorsalis of the laterosphenoid, (pd) dorsal process of the prootic, (Syn) synotic, (VI) foramen nervi abducentis, and (XII) foramen nervi hyppoglossi. Scale bar, 1 cm.

stapes and, judging from its perichondral surfaces, was connected to it by a cartilaginous layer. In Sauroctonus and Suchogorgon, the anterodorsal margin of the impressio basicapsularis has a special area, on which the tuberculum prooticalis ossis stapedis rests (Ivakhnenko, 2005b, p. 427). Biarmosuchus shows a very similar structure of this region (Fig. 32b); however, its area for the tuberculum prooticalis is on the base of the processus paroccipitalis; this enabled the stapes to perform fore-and-aft wobbling movements. Somewhat anterior to this area, the margin of the oval fenestra is elevated and the lower surface of the paroccipital process has a superficial groove (sinus paroccipitalis: Fig. 32b, sp) forming a deep pocket at the base of the process. A very similar structure was described in Suchogorgon and was interpreted as a structure related to the perilymphatic duct (Ivakhnenko, 2005b, p. 434). Unfortunately, the stapes (= columella auris) is only known in a few taxa. In the majority of taxa under study (Nikkasaurus, Alrausuchus, Titanophoneus, Ulemosaurus, Estemmenosuchus, Suminia, all Dicynodontida), the stapes has an expanded base (basis stapedis), which fills the impressio basicapsularis with a very small gap and completely enters the oval fenestra. In Suchogorgon and Biarmosuchus (Fig. 32c), the stapedial base is

distinctly smaller in diameter than the oval fenestra. The anterolateral margin of the base in Suchogorgon (Ivakhnenko, 2005b, text-fig. 29) and the posterolateral margin in Biarmosuchus narrow somewhat and projects medially (tuberculum prooticalis: Fig. 32c, tst). This margin rests on the prootic margin of the impressio basicapsularis; the rest of the bone base does not come in contact with the margins of the impressio basicapsularis, that is, the auditory ossicle was mobile. In Suchogorgon, the stapedial base has a semicircular incisure (alveus rotundus), which forms the lateral border of the foramen rotundum of the perilymphatic duct. In Biarmosuchus (Fig. 32c), the respective margin of the base is slightly depressed, although lacks an incisure; the incisure is also absent in the other taxa investigated. Initially, the stapes of Dinomorpha probably had a bicrural structure, i.e., the stapedial base was connected by narrow crurae to the caput stapedis, and a large oval foramen was located between the crurae. In Nikkasaurus (Fig. 39a), the anteroventral buttress (crus anterius) is massive and curved, whereas the posterodorsal buttress (crus posterius) was very narrow. Only its distal part and tubercle at the base are preserved; however, it remains uncertain whether the buttress was reduced and PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

923

(a)

Sq (b) Po cfl pd cfl fal pd pb Ex

cc clat

cant

Qu

tv

Fig. 37. Periotic regions, sagittal section: (a) Suminia getmanovi Ivachnenko, 1994, specimen PIN, no. 2212/62, braincase, right lateral view; and (b) Delectosaurus berezhanensis Kurkin, 2001, holotype PIN, no. 1536/2, periotic, left lateral view. Designations: (cant) canalis semicircularis anterior, (cc) crus commune, (cfl) canalis floccularis, (clat) canalis semicircularis lateralis, (Ex) exoccipital, (fal) facies laterosphenoidalis of the prootic, (pb) basipterygoid process of the basisphenoid, (pd) dorsal process of the prootic, (Po) postorbital, (Qu) quadrate, (Sq) squamosal, and (tv) tuba vestibuli. Scale bars: (a) 0.5 cm and (b) 1 cm.

replaced by a ligament or damaged during burial. The bicrural structure of the auditory ossicles is retained in Alrausuchus (Fig. 39b), Biarmosuchus (Fig. 32c), Estemmenosuchus (Fig. 15a), and Suchogorgon. In many Dinocephalia (Ulemosaurus, Fig. 14a; Titanophoneus, Fig. 39c), the crus posterius is completely reduced; it is only represented by a small tubercle at the base and a smooth depression, a margin of the intercrural fenestra, which is cut into the surface of the crus anterius. The same structure is characteristic of the auditory ossicles of all Anomodontia examined in this respect (Suminia, Fig. 13b; Australobarbarus, Fig. 39d; Dicynodon amalitzkii). PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

The structure of the distal part (caput stapedis) particularly widely varies in different groups; this is connected with different relationships of the stapes and lamina tenticulata of the quadrate. In the majority of taxa, this part of the stapes is anterocaudally flattened and dorsoventrally expanded. Nikkasaurus (Fig. 39a) lacks a distinct expansion and its very weakly convex distal surface (facies quadrati) adjoins the wide concave surface of the sulcus stapedialis. In Suchogorgon, the plate of the caput stapedis expands mostly downwards, and its flat facies quadrati adjoins the lamina tenticulata of the quadrate (Ivakhnenko, 2005b, textfig. 7). In Ulemosaurus (Fig. 14a), Titanophoneus (Fig. 39c), and Estemmenosuchus (Fig. 15a), the caput

924

IVAKHNENKO (a) (b) cc

cant

cfl

cc cpos

oaa

ap

aa oap

oal

ul

clat

sc

vs

cc

(c) pac

ccl aa ap ut sc vs

fo

ars

S

dp

Fig. 38. Regions of labyrinths of the inner ear: (a) Syodon efremovi (Orlov, 1940), specimen PIN, no. 157/1045, right bony labyrinth open; (b) Alrausuchus tagax (Ivachnenko, 1990), reconstruction of the right membrane labyrinth and cavity of the inner ear; and (c) Delectosaurus berezhanensis Kurkin, 2001, reconstruction of the right membrane labyrinth and cavity of the inner ear. Designations: (aa) ampulla canalis semicircularis anterior, (ap) ampulla canalis semicircularis posterior, (ars) alveus rotundus ossis stapedis, (cant) canalis semicircularis anterior, (cc) crus commune, (ccl) foramen crus canalis semicircularis lateralis, (cfl) canalis floccularis, (clat) canalis semicircularis lateralis, (cpos) canalis semicircularis posterior, (dp) ductus perylimphaticus, (fo) foramen ovale, (oaa) orificium ampullae canalis semicircularis anterior, (oal) orificium ampullae canalis semicircularis lateralis, (oap) orificium ampullae canalis semicircularis posterior, (pac) porus acusticus, (sc) sacculum, (ut) utriculum, and (vs) vestibulum. Scale bar, 0.5 cm.

stapedis expands mostly upwards and continues in a narrow flat band that extends under the lamina postoccipitalis; the facies quadrati, which has a pronounced parachondral surface, does not adjoin the lamina tenticulata, but comes to the sulcus stapedialis. The anterior plane of a slightly curved plate of the caput stapedis adjoins the wall of this groove, and the facies quadrati occupies a postquadrate position. In Suminia, the caput stapedis shows an interesting structure (Fig. 13b), it partially rests on the quadrate, although its posterior

margin expands and comes out to the sulcus stapedialis, but does not enter it. The auditory ossicles of other primitive Anomodontia probably had a similar structure, at least, the stapes of Kawingasaurus was figured similarly (Cox, 1972, text-fig. 2A). In the dicynodonts investigated (Dicynodon amalitzkii, Australobarbarus kotelnitchi: Fig. 39d), the caput stapedis expands only slightly, and its circular in cross section distal margin enters a narrow sulcus stapedialis. The stapes of Alrausuchus is very curiously in structure (Fig. 39b). It PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

925

(b) cra (a)

Ept crp Sq bst

bst

cra

QQJ (c)

Sq

T

poc Sq

Pt

cra

T

(d)

poc

Pt

tv

QQJ

Fig. 39. Auditory regions: (a) Nikkasaurus tatarinovi Ivachnenko, 2000, specimen PIN, no. 162/31, right stapes, anterior view; (b) Alrausuchus tagax (Ivachnenko, 1990), holotype PIN, no. 3706/10, part of the left occipital surface, posterior view; (c) Titanophoneus potens Efremov, 1938, lectotype PIN, no. 157/1, right stapes, posterior view; and (d) Australobarbarus kotelnitshi Kurkin, 2000, holotype PIN, no. 4678/2, left stapes, posterior view. Designations: (bst) basis stapedis, (cra) crus anterius ossis stapedis, (crp) crus posterius ossis stapedis, (Ept) epipterygoid, (poc) paroccipital process, (Pt) pterygoid, (QQJ) quadrate–quadratojugal complex, (Sq) squamosal, (T) tabular, and (tv) tuba vestibuli. Scale bars: (a) 0.2 cm, (b, d) 1 cm; and (c) 2 cm.

has a dorsally expanded caput stapedis, while its lower part is elongated, rounded, enters a narrow sulcus stapedialis, with a postquadrate position of the distal surface, and the ascending plate is thin, narrow, and does not adjoin the margin of the quadrate. 7. Auditory Structures The auditory apparatus of primitive Theromorpha is poorly understood and has not been reconstructed. All researchers in general agree that the mammalian audiPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

tory apparatus developed from the structures connected with the frame of the periangular cavity and the bones of the quadratum–articulare joint of primitive taxa (see De Beer, 1937; Olson, 1944; Parrington, 1967; Allin, 1975; Tatarinov, 1976; Crompton and Jenkins, 1979; Kermack and Mussett, 1983; Luo and Crompton, 1994; Ivakhnenko, 2003c; and others). Therefore, it is natural to propose that the periangular cavity was initially related to the primitive sound-conducting system. The hypothesis that the cavity initially contained a portion of the pterygoid musculature (Watson, 1953; Par-

926

IVAKHNENKO

rington, 1955; Crompton, 1958; Kemp, 1972; De Mar and Barghusen, 1972; Barghusen, 1972, 1973; etc.) or the muscles moving the jaw (Kemp, 1982) is unacceptable (Ivakhnenko, 2003b, 2003c). This hypothesis contradicts the large area of the cavity combined with a very small width, the complex configuration of its interior, and the fact that it is covered externally by a bony wing (ala angularis); these features are inconsistent with the structure of a cavity containing muscles. The bony wing frequently has scalloped lower and caudal margins; this strongly suggests that it was interlaced with dense connective tissue. In addition, the bony wing has a distinct incisure in the upper margin (recessus alae, Fig. 25a), which opens the upper part of the cavity obliquely dorsally and posteriorly. Therefore, it has been proposed that the cavity was initially preformed by the infradental diverticulum of the seismosensory canal of aquatic ancestors (Ivakhnenko, 2003b, 2003c), since only labyrinthodonts have a cavity located in the same position on the surface of the angular bone and containing a sac of a sensory organ, with a foramen open externally. During the transition to terrestrial mode of life, the cavity was retained, changing its function, and was covered externally by a bony outgrowth of the anterior margin (ala angularis), which protected the sac from deformation. In the course of evolution, the sac turned into the mammalian tympanum, forming the double-layer pars tensa in the fork of the annulus tympani, which was formed by remains of the angular bone and angular wing, and the single-layer pars flaccida in place of the recessus alae. The only acceptable assumption concerning the function of this structure is participation in the soundconducting mechanism; however, the mode of functioning remains uncertain (Ivakhnenko, 2003c, pp. 354, 355). Extending available data on the structure of the sound-conducting system in the most primitive tetrapods, it is possible to conclude that conduction of sound vibrations was initially performed through dense tissues in place of the reduced operculum and, then, through a more or less changed hyomandibular and through the oval fenestra of the auditory capsule to the perilymph of the vestibulum (Ivakhnenko, 2003c, p. 349). In fact, all known tetrapod groups use this apparatus, which is improved following various trends. The basic trend is connected with the formation of one or another variant of the tympanic cavity and tympanum. This is observed in anurans, the tympanic cavity of which is formed of a rudiment of the seismosensory placode (probably a preopercular diverticulum). The cavity of higher Parareptilia was probably formed similarly, whereas that of Lepidosauria developed through the anterior curvature of the quadrate; in Archosauromorpha, it was formed through the displacement of the cleidomastoid musculature by a special flank of the paroccipital process. Therefore, it is not surprising that the cavity of mammal ancestors could have been formed ultimately of the material of the infradental diverticulum. A unique feature is the preservation of the

cavity in primitive Theromorpha on the lower jaw; initially, it was relatively far from the hyomandibular, which transformed into the stapes. The recessus alae was probably connected with the external auditory meatus, and the periangular cavity was an efficient resonator amplifying the fluctuations (see Allin, 1986). The small size of the cavity in the only large representative of Sphenacomorpha investigated in this respect (compare Romer and Price, 1940, pl. 22F, Dimetrodon grandis) probably results from the secondary reduction, because even in the most primitive Dinomorpha, the cavity is relatively large. In theory, in the basal groups of Theromorpha, as in all other primitive tetrapods, the distal end of the stapes, which is derived from the hyomandibular, would have been positioned morphologically posterior to the quadrate. In Sphenacomorpha examined in this respect (see, for example, Dimetrodon: Romer and Price, 1940, pl. 14), the stapes is massive, has a relatively small foramen, and its distal margin enters the sulcus stapedialis of the quadrate, and the caput stapedis passes posterior to the quadrate. In Dinomorpha, the impossibility to fix the caput stapedis in a posterior position relative to the quadrate was probably initially connected with the primary streptostyly; however, a similar situation was successfully overcome in the evolution of Lepidosauria. The fact that, in primitive Dinomorpha (for example, Nikkasauridae), the distal head of the stapes rests on the medial surface of the quadrate is an undoubted apomorphy that probably means that, at this evolutionary stage, the use of the resonator periangular cavity was more efficient than the creation of any postquadrate structure. As mentioned above, the periangular cavity (fissura periangularis) always consists of two parts, the upper part (fissura externa) opening in the recessus alae and the lower part (fissura interna), which are separated from each other by the crista alata (Figs. 25a, 25b). The periangular cavity was probably much greater in volume than the space outlined by the ala angularis on the lateral surface of the jaw. The lower margin of the plate is always lowered much below the jaw edge; the fissura interna undoubtedly continues onto the medial surface of the angular bone, occupying the fovea praearticularis and reaching the sulcus angularis superior at the border between the surangular and prearticular bones. The medial part of the cavity is separated by the crista interlaminata from the part of the internal surface of the external crest (lamina reflexa), where the pterygoid muscles could have been inserted. It is possible that the fovea praearticularis is related to the mandibular (lower) diverticulum of the mammalian tympanic cavity (for discussion on this problem, see Tatarinov, 1976, p. 91). Comparative analysis of different dinomorph groups shows that the structures presumably connected with the air sound conduction varied widely. I believe that the long-term discussion about the position of the tympanum of theromorphs in the postquadrate area or the area of the angular wing (see Romer and Price, 1940; Westoll, 1943, 1945; Watson, 1948, 1951, 1954; PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

Parrington, 1949; Tumarkin, 1955; Cox, 1959; Hopson, 1966; Allin, 1975; Luo and Crompton, 1994; and many others) and the wide variety of opinions on this problem are mostly connected with the structure of this region in the particular group studied by the researchers. For this reason I have refused the opportunity to reconstruct the postquadrate auditory structures in all theromorphs (Ivakhnenko, 2005b, pp. 442, 443); this was based mostly on the study of Biarmosuchus and Gorgonopidae. In Dinomorpha, two basic designs, i.e., bicrural– quadrate and monocrural–postquadrate, are recognized. The two designs did not develop in parallel; on the contrary, the second was undoubtedly derived from the first and appeared independently in different lineages of Dinomorpha; therefore, different groups include transitional taxa, which display various combinations of characters of both designs. Primitive groups with a streptostylic QQJ-complex and groups that retained and developed the streptostyly had a design designated as the bicrural–quadrate sound-conducting system. This design is characterized by the bicrural stapes, in which the basis stapedis is connected to the caput stapedis by two narrow crurae, the anteroventral crus anterius, and the posterodorsal crus posterius. The narrow, usually more or less vertically expanded caput stapedis rests on the wide lamina tenticulata of the quadrate; this retains contact between the quadrate and stapes, as the streptostylic quadrate moves. This design is characteristic of the Nikkasauridae and Eotitanosuchidae. It is usually associated with a thin, undulating ala angularis and a crest (crista posterior) on the posterolateral margin of the squamosal, which borders laterally a narrow cavity posterior to the quadrate. Judging from these characters, a similar structure of the auditory region was in the Inostranceviidae, Phthinosuchidae, Rubidgeidae, and Niaftasuchidae, although the stapes of these groups has not been examined. A particular trend in the evolution of the bicrural– quadrate design is displayed by Suchogorgon, Sauroctonus, and probably all Gorgonopidae. In this group (Ivakhnenko, 2005b, p. 429), the upper part of a very thin ala angularis adjoins the surface of the angular; therefore, the recessus alae is very weakly developed. A broad incisure is located between the posterior margin of the ala angularis and well-developed, lowered, and anteriorly curved in the lower part retroarticular process of the articular bone. This incisure obviously contained a connective tissue plate, which was probably formed as a result of the reduction of the lining of the periangular cavity. In this case, the stapes remains bicrural, but becomes narrow and long. The bicrural stapes is also observed in Alrausuchidae; however, this group has lost streptostyly. Therefore, the stapes, which still retains a dorsally expanding caput stapedis (although lacking contact with the quadrate), acquires in the lower part a small projection, which enters a narrow sulcus stapediPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

927

alis and approaches a postquadrate position (bicrural– postquadrate design). A typical monocrural–postquadrate design is observed in taxa that have lost streptostyly, such as Syodon, Titanophoneus, Ulemosaurus, and Dicynodontia examined (Pristerodontidae, Dicynodontidae). In these groups, the stapes has lost the posterodorsal buttress (crus posterius), becoming monocrural; the caput stapedis is widened to a varying extent and enters a special depression (sulcus stapedialis) on the caudal surface of the quadrate. In these groups, this design is associated with a relatively thick wing without a longitudinal crest on the lateral surface and the absence of crista posterior on the posterolateral margin of the squamosal. Therefore, the postquadrate region is open laterally and bordered medially by the lamina postoccipitalis, a special plate of the lateral margin of the paroccipital process. The occipital surface of the squamosal has a fossa elliptica. The secondary character of the postquadrate position of the stapes in Dinomorpha is supported by the preservation in many groups of structural features connected with the primary design. In particular, in Suminia, the crus posterius is reduced and the stapes becomes monocrural, and the widened lower part of the caput stapedis extends into the sulcus stapedialis. The upper part of the caput stapedis retains the shape of a thin flat process, but lacks contact with the medial surface of the quadrate. The postquadrate cavity is already laterally open (crista posterior is absent); however, the lamina postoccipitalis and fossa elliptica are very poorly developed. The ala angularis remains thin, with a horizontal crest. The design of the Ulemicidae is very similar, in which the structure of the stapes is not known. On the contrary, Estemmenosuchus has a massive and thick ala angularis, with a smooth lateral surface, well developed fossa elliptica and lamina postoccipitalis, laterally open postquadrate cavity, and a wide sulcus stapedialis on the caudal surface of the quadrate, which contained a wide caput stapedis, although the stapes is bicrural (bicrural–postquadrate design). The functioning mode of all the designs described remains poorly understood. The initial bicrural–quadrate design is particularly difficult to interpret. In some taxa (such as Alrausuchus and Biarmosuchus), the stapes was mobile with certainty, because the stapedial base has a special tuberculum prooticalis, which rests on the fossa at the base of the paroccipital process, and the paroccipital sinus of the perilymphatic duct is well developed. However, this position of the tubercle allows only oblique fore-and-aft wobbling movements of the stapes along the plane of the lamina tenticulata of the quadrate. At the same time, the body of the stapes is strongly widened, forming a vertical frame of narrow crurae and a large intercrural foramen; the plane of the frame is positioned vertically, almost perpendicular to the sagittal plane of the skull. It is hardly probable that all these structural characters correspond to direct sound conduction from the periangular cavity through

928

IVAKHNENKO

the massive and immobile (in the intracranial direction) quadrate and articular bones. The intercrural foramen was probably covered by a membrane, and vibration was perceived by the entire frame of the stapes. However, in this case, vibrations probably came mostly from the posterior and somewhat lateral sides. This region has a relatively small postquadrate cavity between the quadratum body and the cleidomastoid muscles, which is displaced posteriorly by a welldeveloped mastoid process. However, the postquadrate cavity was closed laterally by the crista posterior of the squamosal. A single foramen that opens in this cavity is the foramen paraquadratum, which expands widely posteriorly. Anteriorly, the paraquadrate canal opens in a deep depression on the anterior surface of the quadrate, which is located just above the caudal end of the sulcus angularis superior, which continues the fovea praearticularis distinctly connected with the fissura interna of the periangular cavity, as indicated above. In this place, Biarmosuchus has a broad inflated sulcus angularis superior (sinus angularis); the strong development of this structure is evident from the fact that it corresponds to an eminence (eminentia angularis) on the lateral surface of the angular bone, just anterior to the periangular cavity. In this case, it is possible to propose that the air wave entering the foramen of the recessus alae amplified in the periangular cavity, passed through the sulcus angularis superior in the foramen paraquadratum, and entered the postquadrate cavity, causing fluctuations of the stapedial frame. I believe this is too complex a reconstruction; however, it is hardly possible to propose another explanation for the function of the bicrural–quadrate design. Apparently, the auditory region of Gorgonopidae could have worked much simpler. In this case, the entire thin and flat postdentary part of the lower jaw, which formed a mobile articulation with the dentary, probably functioned as a huge “tympanum” (see Parrington, 1955; Kemp, 1969; Allin, 1975, 1986; Tatarinov, 2000; Ivakhnenko, 2003b, 2003c). A high plate of the caput stapedis of a relatively thin stapes adjoins the lamina tenticulata of the quadrate, and the widened base (basis stapedis) is somewhat smaller in diameter than the foramen ovale and has a special tubercle (tuberculum prooticalis), which rests on the impressio basicapsularis on the margin of the fenestra; this allowed small-amplitude wobbling movements of the stapes mostly in the intracranial direction. Thus, these taxa have a perilymphatic duct, fenestra rotunda, and a well-developed sinus paroccipitalis on the lower surface of the paroccipital process of the opisthotic. In this design, the stapes is relatively narrow, elongated, with a narrow intercrural foramen; consequently, the above assumption of the use of this bone as a frame for the tympanum is inadmissible. The dorsoventrally expanded distal area of the flattened caput stapedis freely adjoins the broad lamina tenticulata of the quadrate. The fenestra rotunda is formed by a semicircular incisure in the basis

stapedis and the sinus paroccipitalis on the lower surface of the paroccipital process of the opisthotic. It is possible that the monocrural–postquadrate design was an attempt to create a postquadrate tympanum, which is typical of the majority of tetrapods. The postquadrate region is open laterally and increased by the fossa elliptica, and a wide sulcus stapedialis and an overhanging margin of the paroccipital process (lamina postoccipitalis) form a tympanic cavity of a sort. However, in the majority of known taxa, the basis stapedis is somewhat wider than the opening in the foramen ovale (porus acusticus), and the stapes was immobile. The distal margin of the caput stapedis is flat, with a rough surface, the character of which suggests contact with connective tissue rather than the presence of a cartilaginous continuation. Probably, there was only a relatively small connective tissue disk in this area. It is hardly probable that this system was efficient; however, in turtles, which show a very similar structure of this region, with an even smaller relative area of the cross section of the sulcus stapedialis, sound conduction through the stapes plays a significant role (Wever and Vernon, 1956). It is almost indisputable that the soundconducting system of Dinomorpha included the resonator (?) periangular cavity, because the thickening of the ala angularis (see above) is accompanied by an increase in the volume of the cavity, and, in higher Dicynodontia (for example, Idelesaurus and Dicynodon), the angular wing gently deviates laterally, and the cavity becomes a wide pocket. However, in this case, connection between the tympanic and periangular cavities becomes completely uncertain, because, for example, in Deuterosaurus, the paraquadrate canal seems closed; the sulcus angularis superior of Syodon is well pronounced, but lacks caudal expansions; and in Ulemosaurus, it almost disappears as a separate structure, fusing with the fovea praearticularis. It is noteworthy that, as this design is formed, the displacement of the distal end of the stapes to the sulcus stapedialis means that it occupies a lower position. In Dinocephalia, this is compensated by the strongly expressed klinorhiny of the braincase, while, in the nonklinorhinal Dicynodontidae, the tuba vestibule is formed and the margins of the foramen ovale expand laterally and downwards, displacing the stapes downwards. It is possible that, in some taxa, the region of the incisure in the upper margin of the ala angularis (recessus alae) included some structures related to the external auditory meatus. A pronounced groove always extends from the incisure somewhat dorsally and posteriorly, and this region is almost always covered dorsally and somewhat anteriorly by a thickened lower margin of the zygomatic arch. This margin is thickened even in taxa that lack distinct pachyostotic expansions in other regions of the skull (Nikkasaurus, Alrausuchus, Biarmosuchus, etc.). The thickening is frequently rugose pachyostotic externally and always smooth from within, above the groove. If the zygomatic arch deviates from the lower jaw surface (Titanophoneus, EstemPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

929

(a) (f)

(c)

lc ccn (b) (d)

(e)

Fig. 40. Teeth: (a–d) Reiszia gubini Ivachnenko, 2000, holotype PIN, no. 162/32: (a) third right upper incisor, labial view; (b) third left lower incisor, lingual view; (c) upper canine-shaped tooth, labial view; and (d) 12th left lower tooth, lingual view; (e) Reiszia tippula Ivachnenko, 2000, holotype PIN, no. 4541/2, right cheek tooth, lingual view; and (f) Nikkasaurus tatarinovi Ivachnenko, 2000, specimen PIN, no. 162/31, 11th right upper tooth, lingual view. Designations: (ccn) centrocone and (lc) lateroconules. Scale bar, 0.1 cm.

menosuchus mirabilis), the region is covered anteriorly and dorsally with a special rugose bony ridge on the surface of the angular. In Anomodontia, the zygomatic arch of which curves upwards, the region of the recessus alae is usually covered from above by a more or less developed longitudinal crest on the surangular and on the posterior margin of the dentary. It is probable that the groove contained a soft-tissue external auditory meatus, which probably opened posterior to the region of the quadratum–articulare joint. In the Gorgonopidae, which lack this groove because the periangular cavity is flattened, a thin vertical crest (spina angularis) is positioned just anterior to the periangular cavity; this crest could have been related to certain structures of the external ear (Ivakhnenko, 2005b, p. 442). 8. Dentition The teeth of various Dinomorpha consist of a more or less expanded crown (corona dentis) covered by enamel, a distinct neck (collum dentis) lacking enamel, and a long root (radix dentis) tapering towards the apex, with a small foramen apicis at the tip. An exception is the incisors of Ulemicidae and tusks of Dicynodontidae, with constant growth and cylindrical root open at the apex. The canalis radicis passes inside from the foramen apicis and forms a cavity (cavum dentis) inside the tooth. The relative size of the cavum dentis decreases during morphogenesis, because the crown PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

grows by addition from inside of subconical layers, which are thickest at the apex of the crown and narrow towards the root. These concentric layers are usually distinct in almost all broken teeth. Of the groups investigated, the Nikkasauridae have the simplest teeth (Fig. 40). The anterior jaw teeth have a wide neck and a more or less extended and posteriorly curved crown (centroconic type, Figs. 40a–40c). Posteriorly, the cheek teeth become straighter, somewhat flattened in linguolateral direction, with sharp crests on the anterior and posterior edges. The crown is widened relative to the collum dentis. The enamel surface usually has a characteristic sculpture of thin undulating ridges with varicoid thickenings stretching from the apex. As these thickenings crossed the cutting borders, the primary tubercles could have been formed to increase the durability and cutting properties of the borders. Enlarged tubercles first appear on the posterior border and, then, develop on the anterior border (Figs. 40d–40f); subsequently, both borders become regularly serrated throughout the extent (lateroconule type). This dental structure shows the morphology initial to all Dinomorpha. All structural tooth variants in various groups result from plastic deformations of this initial type. The crown of the initial tooth is a cone flattened in linguolateral direction. Its apex is designated as centrocone (Fig. 40d, ccn), because homology with mammalian cusps is uncertain. The labial surface of

930

IVAKHNENKO (b) (a) (c)

Fig. 41. Teeth of Niaftasuchus zekkeli Ivachnenko, 1990: (a) specimen PIN, no. 4543/20, anterior incisors of the right dentary, lingual view; (b) specimen PIN, no. 4660/31, isolated first lower incisor, lingual view; and (c) specimen PIN, no. 3717/39, isolated cheek tooth, lingual view. Scale bars: (a) 0.5 cm and (b, c) 0.1 cm.

(c)

(a)

(d) (b)

(f)

(e)

Fig. 42. Teeth: (a) Ustia atra Ivachnenko, 2003, holotype PIN, no. 4548/155, fourth right postcanine, lingual view; (b) Rhopalodon sp., specimen SGU, no. 104B/2050, third right lower postcanine; (c) Phthinosaurus borissiaki Efremov, 1940, holotype PIN, no. 164/7, posteriormost tooth of the right dentary, replacement crown; (d, e) Phthinosaurus sp., specimen PIN, no. 4539/1: (d) replacement crown of a cheek tooth and (e) cheek tooth, lingual view; and (f) Rhopalodon (?) sp., specimen PIN, no. 270/2, upper (?) canine, lingual view. Scale bars: (a–e) 0.2 cm and (f) 0.5 cm. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA) (b)

931

(c)

(f)

(e)

(a)

(d)

Fig. 43. Teeth: (a–c) Parabradysaurus silantjevi Ivachnenko, 1995, (a) specimen PIN, no. 4416/36, right upper canine, lingual view; (b) specimen PIN, no. 4416/9, isolated lower incisor, lingual view; and (c) specimen PIN, no. 4416/4, isolated cheek tooth, lingual view; (d–f) Estemmenosuchus mirabilis Tchudinov, 1968: (d) specimen PIN, no. 1758/336, isolated cheek tooth, lingual view; and (e, f) holotype PIN, no. 1758/6: (e) cheek tooth, lingual view; and (f) right lower incisor, lingual view. Scale bars: (a) 1 cm and (b–f) 0.5 cm.

this cone is slightly convex and has narrow anterior and posterior borders consisting of varying number of tubercles (lateroconules, Fig. 40f, lc). This is probably an optimal tooth design, which is observed in all jaw teeth. The upper and lower jaw teeth of Dinomorpha almost always display symmetry relative to each other; therefore, descriptions of the upper and lower jaw teeth are identical. Thus, the working surface of teeth is designated below as occlusal rather than upper or lower, although a more or less pronounced occlusion of teeth antagonists of Dinomorpha is only observed in incisor of some groups and canines of predators. Hence, it is possible to apply the term distal to designate the apex of canine-shaped teeth. Structural changes in teeth depend on the position in the jaw and, hence, on the main function. As mentioned above using the example of Nikkasauridae, the primary division is caused by the differentiation into anterior teeth, mostly grasping (incisors), and cheek teeth, mostly treating food objects. In the evolution of Dinomorpha, the cheek teeth, which were exposed to miniPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

mum stresses, changed relatively slowly; the teeth of the incisive region (located in the premaxilla and the anterior part of the maxillary and dentary), which underwent a greater stress, changed much more rapidly. It is noteworthy that constructional novelties improving the work of the most loaded teeth of the incisive region often extend to all teeth, including cheek teeth, where they play a minor role. The simplest variations in the primitive type of cheek teeth concern differences in the crown proportions, the height-to-width ratio, and in the number and relative size of the lateroconules. Two simplest forms are recognized, i.e., elongated piercing–cutting (termed long-leaf to designate a particular type rather than to describe it), with many small lateroconules (Figs. 40e, 42a, 42b, 49a), and restricting (food objects in the mouth cavity) short-leaf (designating a tooth of the system where the cheek teeth form lateral borders of the mouth cavity; this system may be designated briefly as a fence structure), with a relatively small number of increased lateroconules (Figs. 41c, 43f, 43d, 46e, 55a).

932

IVAKHNENKO

(d) (a)

(e)

(b)

(c)

Fig. 44. Teeth: (a) Kamagorgon ulanovi Tatarinov, 1999, holotype PIN, no. 4312/1, third left upper tooth, lingual view; (b–e) Inostrancevia sp., (b) specimen PIN, no. 2005/1775, isolated left lower incisor, lingual view; (c–e) specimen PIN, no. 2005/1757, right upper canine: (c) labial view, (d) cross section, and (e) fragment of cutting border of the lower canine, magnified. Scale bar, 1 cm.

An increase in specialization of these teeth resulted in the development of carnassials, i.e., cutting teeth with a strengthened base, posteriorly curved apex, and long sharp serrated border (Figs. 44a, 48c, 50a, 50d), and short crushing teeth shaped like rounded tubercles almost losing cutting crests (Figs. 47c, 47d, 51h). In the last case, the number of teeth in a row sometimes increased, providing additional crushing surface (Syodon: Fig. 27a; Ulemica). The loss of the basic cutting function of the cutting teeth and preservation of the holding function (Gorgonopidae, Figs. 45c, 45d; and Inostrancevia, without lower jaw antagonists) resulted in simplification, the teeth became more cylindrical, with weak borders. In Gorgonopidae and Inostranceviidae, the teeth of the incisive region have changed to the least extent (Figs. 44b, 45a). The lingual surface of these teeth is convex, so that they are almost circular in cross section and the cutting crests are pulled somewhat closer lin-

gually, providing a united semicircular cutting surface of the incisor row. The incisors of Niaftasuchidae (Figs. 41a, 41b), Venyukovia (Figs. 53a, 53b), Ulemica (Fig. 53e), and Suminia (Figs. 54a, 54b) are also circular in cross section, but have a somewhat concave lingual surface; they are inclined anteriorly and, because of rubbing against the incisor–antagonist, become dolabriform, with a sharp flattened apex (see, for example, Fig. 53c). The flattened crown is reinforced by the increased varicoid crests (Figs. 41a, 41b, 47a, 50c, 51c) or only one increased middle crest, which forms a longitudinal counterfort (Figs. 43b, 43c). Marginal lateroconules have disappeared or become small, demolished by wear. In these incisors, the greatest stress probably falls on the crown base, at the point of transition to the somewhat narrowing collum dentis. Therefore, this region is reinforced by the thickened crown base on the lingual side or even both sides (Ulemica, Suminia). The crown expands linguolabially, the centrocone base projects PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

(c)

933

(d)

(b)

(a)

Fig. 45. Teeth: (a–c) Suchogorgon golubevi Tatarinov, 2000, (a) specimen PIN, no. 4548/158, first left mandibular incisor, lingual view; (b) specimen PIN, no. 4548/58, isolated left maxillary canine, lingual view; and (c) specimen PIN, no. 4548/161, isolated postcanine, lingual view; (d) Sauroctonus progressus (Hartmann-Weinberg, 1938), specimen PIN, no. 156/60, isolated postcanine, lingual view. Scale bar, 0.5 cm.

labially, and the lower edges of the cutting crests continue onto the thickenings of the crown base. As a result, a cingulum-like structure is formed at the crown base; this is most pronounced in Dinocephalia (Fig. 46c). In subsequent evolution of many groups, the structure strengthening the tooth base extends onto the cheek teeth (Figs. 46e, 52d). This dental structure is assigned to a particular type (cinguloid type). Depending on the function, the teeth of this type vary in the relative height of the centrocone, the width and development of the cinguloid. It is noteworthy that, in Estemmenosuchida, particular lower lateroconules extend onto the lingual surface of the tooth (e.g., in Phthinosaurus: Figs. 42c– 42e, this is also observed in some jaw teeth of Estemmenosuchus); however, a cinguloid is not formed, although their cheek teeth are similar in appearance to those of Dinocephalia. In primitive Anomodontia, serrated cinguloids are formed on both lingual and labial sides of the crown. They are particularly strongly developed in Suminia, but preserved only on the crowns of very young individuals (Figs. 54c, 54d); as a tooth erupted from the alveolus, these structures rapidly disappeared because of wear. Therefore, during the ontogeny, the shape of the replacement crown of Suminia changed, becoming wider and somewhat flattened longitudinally on the occlusal surface, as though accomPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

modating to the plane of wear (Figs. 54e, 54f). At the same time, the cinguloids decrease considerably in size on both lingual and labial sides. In primitive Anomodontia (Ulemica, Suminia), the crown base of incisors is reinforced on both lingual and labial sides (Figs. 53c, 54a, 54b), forming a double V-shaped cinguloid on cheek teeth. It is evident that the lateroconules closest to the collum dentis participate in the formation of the lingual cinguloid; however, the mechanism of the formation of the labial cinguloid remains uncertain. In some groups of Dinocephalia, the labial displacement of the centrocone and the expansion of the cinguloid are well-pronounced, such that the occlusal surface of the anterior incisor becomes flattened and widened oviform (Figs. 46a, 46b), increasing in flatness with wear. At the same time, the lingual side of the tooth narrows, and the edges with lateroconules are displaced closer. This structure is probably characteristic of various groups, since it sometimes occurs as a deviation even in juvenile teeth of Anomodontia (Fig. 45b). It is noteworthy that, in one specimen of Titanophoneus (PIN, no. 157/1), the first upper right incisor is double and is very similar in shape (Fig. 48a) to the incisor of Ulemosaurus (Fig. 46a). At the same time, the normal left incisor of this specimen shows the same shape as the lower incisor (Fig. 48b). The anterior

934

IVAKHNENKO (a)

(e)

(d)

ccn

(b)

(c)

Fig. 46. Teeth of Ulemosaurus svijagensis Riabinin, 1938: (a) specimen PIN, no. 2207/3, isolated right upper incisor, lingual view; (b) specimen PIN, no. 2207/14, isolated replacement incisor, posterolabial view; (c) specimen PIN, no. 157/243, isolated second left lower incisor, posterolabial view; (d) specimen PIN, no. 157/238, isolated right upper canine-shaped tooth, lingual view; and (e) specimen PIN, no. 157/222, isolated cheek tooth, lingual view. Designations: (ccn) centrocone. Scale bar, 1 cm.

incisors of Microurania display a similar structure (Figs. 52a, 52b, 52e), with similar sharply widened crowns, but without traces of rubbing against the opposite incisors. A number of groups of Dinomorpha have acquired increased canine-like teeth in the upper jaw, or even true canines in the upper jaw or both jaws. The increased teeth are formed undoubtedly secondarily through an increase in size of a jaw tooth. This is seen in the increased tooth of a young individual of Ulemica (Fig. 53d), which is identical in shape to usual jaw teeth (in old animals, the increased tooth is heavily worn by rubbing against a rugose surface of the lower jaw, which almost completely changes its initial shape). In

Niaftasuchus, the increased fifth upper tooth becomes similar in structure to its incisor. This tooth has a strongly convex anterior surface and a somewhat concave posterior surface, which has two or three reinforcing longitudinal ridges. The canine-shaped upper tooth of Ulemosaurus (Fig. 46d) differs from the abovedescribed tooth of Niaftasuchus almost only in size and elongation and differs considerably from the neighboring incisors (identical in shape to the lower tooth shown in Fig. 46c). An increase in size and relative length of such a tooth and the development of its antagonist in the lower jaw probably resulted in the formation of “canines” characteristic of a number of other Dinocephalia, for example, Deuterosaurus and TitanoPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA) (a)

(b)

(c)

935

(d)

Fig. 47. Teeth of Deuterosaurus biarmicus Eichwald, 1846, specimen PIN, no. 1954/1: (a) right second lower incisor, lingual view; (b) canine-shaped lower tooth, lingual view; (c) cheek tooth, lingual view; and (d) replacement posteriormost cheek tooth, lingual view. Scale bar, 0.5 cm.

(g) (e) (f) (c)

(a) (d) (b)

Fig. 48. Teeth of Titanophoneus potens Efremov, 1938, lectotype PIN, no. 157/1: (a) first and second right upper incisors; (b) third right lower incisor, laterolingual view; (c) third right postcanine, lingual view; (d) left lower “canine,” lingual view; (e–g) right upper “canine” (e) labial view, (f) cross section, and (g) fragment of the cutting border of the lower “canine,” magnified. Scale bar, 1 cm. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

936

IVAKHNENKO (b)

(c)

(a)

(d)

Fig. 49. Teeth of Alrausuchus tagax (Ivachnenko, 1990), specimen PIN, no. 4659/8: (a) posteriormost lower cheek tooth, labial view, (b) right upper canine, labial view; (c) cross section, and (d) fragment of the cutting border of the lower canine, magnified. Scale bar, 0.5 cm.

phoneus (Fig. 48e). These upper “canines” usually retain a characteristic thickening at the point of transition to the collum (it is particularly well pronounced in Deuterosaurus, specimen PIN, no. 2629/1); they are almost circular in cross section (Fig. 48f). The lower “canines” of these groups are even more similar to incisors (Figs. 47b, 48d), differing only in the weaker developed cinguloid. The cutting borders of these “canines” retain small lateroconules, varying in size and space between them (Fig. 48g). The canine-shaped teeth of Microurania (Figs. 52c, 52f) are very similar in structure to its cheek teeth, differing in the extended apex. They have a distinct cinguloid with one or two tubercles on the lateral margins, but lack lateroconules along the high apex. It is possible that, in this case, the entire upper part of the “canine” is an increased centrocone. The upper tusks of Dicynodontia (Figs. 9b, 55d) probably followed a special developmental mode. None of the taxa examined show a trace of lateroconules or cinguloids in definitive or juvenile canines (which are represented by many specimens). Since the teeth of primitive Anomodontia and the cheek teeth of Dicynodontida (if preserved) have lateroconules (Fig. 55a, 55b), it is possible that such a “canine” is

merely an expanded apex (centrocone) with undeveloped rudiments of the lower part, including lateroconules. A juvenile “canine” looks like a very thinwalled, slightly curved cone (Fig. 55c), the apex of which enters a small alveolus. As an animal grew, the alveolus expanded and the “canine” came out, constantly depositing conical layers from within the tooth. As the definitive developmental stage was achieved, the tusk stopped growing in diameter and became cylindrical in shape (Fig. 9b). An unusual canine structure is characteristic of Alrausuchus (Fig. 49b), Biarmosuchus (Figs. 50b, 50f), all Gorgonopidae (Fig. 45b), Phthinosuchidae, and Inostrancevia (Fig. 44c). Both upper and lower canines of identical shape are present. They are very long, flattened considerably laterally, teardrop-shaped in cross section, and narrowed posteriorly (Figs. 44d, 49c). Many lateroconules are located on sharp anterior and posterior borders, forming an even row and having the shape of cutting blades (Figs. 44e, 49d) rather than rounded tubercles (which are present on other teeth). The canines curve abruptly in the shape of sickles; the upper and lower canines form a single cutting system of alternative canines (Ivakhnenko, 2003c, p. 359). The PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA) (f)

(a) (d)

(c)

937

(b)

(e)

Fig. 50. Teeth of Biarmosuchus: (a, b) B. tchudinovi Ivachnenko, 1999, holotype PIN, no. 4309/1: (a) fifth left postcanine, lingual view and (b) left canine, lingual view; (c–f) B. tener Tchudinov, 1960: (c) specimen PIN, no. 1758/334, isolated first left incisor, lingual view; (d) specimen PIN, no. 1758/212, isolated canine of a young individual, (e) specimen PIN, no. 1758/255, right upper postcanine, lingual view; and (f) specimen PIN, no. 1758/255, left lower tooth, lingual view. Scale bars: (a, c–e) 0.5 cm and (b, f) 1 cm.

canines of Alrausuchus and Biarmosuchus are almost straight in the parasagittal plane; in Inostrancevia, they curve slightly lingually at the base and, in the upper half, curve slightly laterally. The same curvature, although somewhat less expressed, is characteristic of canines of all Gorgonopidae under study (Ivakhnenko, 2005b, p. 433). The canines of relatively small Alrausuchidae and Gorgonopidae are small relative to the skull size, thickened, and curved, while the large saber-shaped canines of giant Eotitanosuchidae and Inostranceviidae are straighter and more flattened; this probably reflects similar functioning. At the same time, the canines of Gorgonopidae and Inostranceviidae are wider before the cutting borders; this is probably connected with their origin, since the canines of gorgonopians developed by elongation (accompanied by multiplication of cutting lateroconules) of carnassials widened at the base (such as shown in Fig. 44a) rather than analogous transforPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

mation of a relatively flat long-leaf tooth, with a narrow neck, as in dinocephals (Figs. 50d, 50e). As the function changed, the canines were modified, losing cutting properties and turning into tusks circular in cross section, as is characteristic of Rhopalodontoidea. At the beginning, the marginal borders are retained (the level of Rhopalodon, Fig. 42f), but lose almost completely a cutting serration, becoming irregularly arranged expansions scattered along the border. As the primary varicoid ridges on the tooth surface are reinforced and expanded, the crown becomes distinctly faceted; this probably provides mechanical strengthening of the tusk. Subsequently (at the level of Parabradysaurus, Fig. 43a), the marginal borders show the same structure as the edges of facets; the tusk becomes even more massive and more circular in cross section. In Estemmenosuchus, the tusks are circular in cross section, and even its incisors have lost borders and become

938

IVAKHNENKO (a)

(f)

(b)

(e)

(c) (d)

(h) (g)

Fig. 51. Teeth: (a, b) Microsyodon orlovi Ivachnenko, 1995, holotype PIN, no. 4276/13: (a) right upper “canine,” lingual view; (b) sixth right upper postcanine, lingual view; (c–f) Archaeosyodon praeventor Tchudinov, 1960: (c) specimen PIN, no. 1758/95, right anterior lower incisor, lingual view; (d) specimen PIN, no. 1758/315, isolated upper replacement “canine,” labial view; (e) specimen PIN, no. 1758/328, right lower “canine,” labial view; and (f) specimen PIN, no. 1756/95, fifth right lower postcanine, lingual view; (g, h) Syodon efremovi (Orlov, 1940): (g) specimen PIN, no. 157/2, right upper canine, labial view; and (h) specimen PIN, no. 157/677, seventh left lower postcanine, lingual view. Scale bar, 0.5 cm.

slightly faceted (Fig. 43e). Similar changes probably occurred in Dinocephalia, as the Archaeosyodontidae were formed. In primitive taxa (e.g., Microsyodon, Fig. 51a), the canine is slightly faceted, like that of Rhopalodon, has weak traces of serration on the bor-

ders, and is curved considerably posteriorly. Subsequently, the canine became even more circular in cross section and more strongly curved (Archaeosyodon, Fig. 51d). The lower canine of this genus (Fig. 51e) is short, with a posteriorly curved apex; however, it PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

939

(a) (c) (d)

(b) (f)

(e)

Fig. 52. Teeth of Microurania: (a–d) M. minima Ivachnenko, 1995, holotype PIN, no. 4337/1, (a) first right upper incisor, lingual view; (b) dorsal surface of the first right lower incisor; (c) right lower canine-shaped tooth, lingual view; and (d) right lower cheek tooth, lingual view; (e, f) M. mikia Ivachnenko, 2003: (e) holotype PIN, no. 4538/7, fragment of the left ramus of the lower jaw, and (f) specimen PIN, no. 4538/29, upper canine-shaped tooth, lingual view. Scale bar, 0.1 cm.

retains distinct traces of a cinguloid thickening at the crown base and has much in common with the incisor (Fig. 51c). This trend in the development of the canine is most pronounced in Syodon (Fig. 51g), the canine of which is sharply curved, cylindrical, and lack borders. These changes in the canine are accompanied by transformation of the cheek teeth, which have become shorter, more massive, and adapted for crushing (Figs. 51b, 51f, 51h). In Dinomorpha, regular tooth wear is only observed in special cases, because regular occlusion is very rare. Wear facets usually developed because of relatively unstable arrangement of teeth in the upper and lower PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

jaws and did not improve functional properties of teeth. The facets appeared because of contacts between opposed crowns; they occupy varying positions on the crowns and are not necessary present in all jaw teeth. Such facets are usually present on densely positioned incisors (see, for example, Figs. 41b, 43b), but are infrequent on cheek teeth. Wear traces rarely occur on the cheek teeth, which are small and almost lack occlusion in Niaftasuchus, Ulemosaurus, Rhopalodon, Phthinosaurus, Parabradysaurus, Estemmenosuchus, and Biarmosuchus (Fig. 50d). On the contrary, they are present on almost all cheek teeth of Microurania (Fig. 52d) and Alrausuchus (Fig. 51a), which are large

940

IVAKHNENKO

(e)

(d)

(c) (a)

(b)

Fig. 53. Teeth: (a, b) Venyukovia prima Amalitzky, 1922, lectotype PIN, no. 48/1, (a) tooth row of the left ramus of the lower jaw and (b) first right incisor, labial view; (c, d) Ulemica invisa (Efremov, 1938): (c) specimen PIN, no. 157/1117, isolated first replacement lower incisor, lingual and somewhat lateral view, and (d) specimen PIN, no. 157/989, increased upper tooth of a young individual, labial view; and (e) Ulemica sp., specimen PIN, no. 519/1, right premaxillary, lingual view. Scale bar, 0.5 cm.

relative to the jaw size. No wear facets are observed on the cheek teeth of Gorgonopidae, because their upper and lower teeth do not contact tightly, and on the upper cheek teeth of Inostrancevia because it lacks lower jaw antagonists. Constant and natural tooth wear connected with the kinematics of the jaw system, but not improving the functional properties of teeth, is described for the canines of Suchogorgon (Ivakhnenko, 2005b, p. 433). On the upper canine, the wear facet occupies the apex and ascends along the anterior border. On the posterior border, the wear facet is usually located within its lower quarter. On the lower canine, the wear facets are located at the base of the posterior border and in the upper part of the anterior border. Since direct occlusion of the upper and lower canines does not result in the formation of such wear facets, they are probably formed due to complex relative movements of jaws in a streptostylic skull. In Biarmosuchus and Inostrancevia, which also have a streptostylic skull, similar facets are much weaker or absent; this is probably attributable to the less curved canines of these taxa.

Constant functional wear that undoubtedly improves tooth properties is observed in the anterior incisors of Ulemosaurus (Fig. 46a). The occlusal surface of such an incisor with a very wide cinguloid becomes even wider because of vertical rubbing against the surface of the lower incisor. As a result, the occlusal surface of the upper incisor becomes somewhat depressed, while that of the lower incisor becomes convex; this strongly suggests the absence of horizontal jaw movements. This is corroborated by the constant presence of a narrow vertical wear facet on the posterolingual surface of the canine-shaped tooth (Fig. 46d), which is produced by the apex of the lower tooth. Apparently, the tusk structure of Dicynodontia was also improved, the tusks frequently have very unusual flat wear facets on the lingual surface (Figs. 9b, 55d). This sharpening mode is only attributable to rubbing against the horn cover of the lower jaw. Very similar self-sharpening incisors are observed in a number of taxa (e.g., Venyukovia, Fig. 53b; Ulemica, Fig. 53e; Suminia, Fig. 56a); this is performed by incisors antagonists and results in sharpening both upper and lower incisors. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

(c)

941

(d)

(f) (a)

(b) (e)

Fig. 54. Teeth of Suminia getmanovi Ivachnenko, 1994: (a, b) specimen PIN, no. 2212/82: (a) first right lower incisor, lingual view, and (b) left anterior incisors, labial view; (c, d) specimen PIN, no. 2212/33, replacement cheek tooth of a very young individual: (c) lingual and (d) labial views; (e, f) specimen PIN, no. 2212/18, replacement definitive crown of a cheek tooth: (e) lingual and (f) labial views. Scale bars: (a, b) 0.5 cm and (c, d) 0.1 cm.

The wear of cheek teeth in Venyukovia (Fig. 53a) is probably a special case. The teeth are arranged somewhat irregularly and sharpened by rubbing against the teeth–antagonists; however, wear facets are positioned in different planes; this strongly suggests that the teeth were worn only by vertical jaw movements. Unfortunately, the upper teeth have not been recorded in this taxon. However, in Ulemica, with undoubtedly only vertical jaw movements (the dentary has a depression corresponding to an increased upper tooth), wear facets on the premaxillary incisors are shaped like semicircular depressions (Fig. 53e); this also corresponds to vertical movements. A different picture is observed in the cheek teeth of Suminia, wear facets of which are confined to one longitudinal plane (Fig. 56b). In the first premaxillary incisor, the wear facet is in the shape of a circular depression; in the others, they are flat, located in the same plane of wear (Fig. 56a). These data correspond to longitudinal movements of the lower jaw, PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

which, in the case of extreme protraction, did not extend beyond the anterior incisor. Sometimes, wear considerably distorts teeth; therefore, morphological descriptions and figures in this work are based where possible on the crowns of replacement teeth. In all Dinomorpha, the teeth were replaced according to a standard pattern, an anlage of a replacement tooth is located in a fossa at the posterolingual edge of the root of the replaceable tooth. Initially, the upper part of the crown with the centrocone was formed; subsequently, a complete crown developed, and, after resorption of the base of the replaceable tooth, it dropped out, and a replacement tooth appeared in its place simultaneously with the formation of the root. The teeth were particularly intensely replaced in Suminia. During functioning, its teeth were worn out almost to the collum (Fig. 56c) and replaced by the teeth of the next generation. The teeth were replaced several times during

942

IVAKHNENKO (d) (a)

(b) (c)

Fig. 55. Teeth: (a, b) Australobarbarus sp., specimen PIN, no. 4678/8, lower replacement teeth, lingual view; (c, d) Dicynodontidae gen. indet., isolated tusks: (c) specimen PIN, no. 1538/60, crown of a juvenile, and (d) specimen PIN, no. 1538/61, left definitive tooth, lingual view. Scale bars: (a, b) 0.5 cm and (c, d) 1 cm.

the animal’s life (Fig. 57b); occasionally, up to three tooth generations are present simultaneously. The replacement of canines was also intense in predators, with alternative cutting canines and special capsules for replacement canines (Alrausuchus, Biarmosuchus, Sauroctonus, Suchogorgon: Ivakhnenko, 2005b, p. 444; Inostrancevia). The replacement canine has a very thin-walled crown extending to the collum; as the base of the replaceable canine was resorbed, this crown erupted in the major alveolus. As the root was formed, the canine projected from the alveolus and, due to the presence of a special capsule, where the crown was almost completely formed, this process could have been very rapid. Note that the details of this process remain uncertain and require additional study. In particular, the capsule of the replacement canine of Biarmosuchus is always positioned posterior to the capsule of the major canine, i.e., posterior to the root of the replaceable canine.

As the crown of the major canine was reduced, the crown of the replacement canine would have been located in the alveolus posterior to the remains of the root of the major canine; however, in the cases where the position of the replacement crown is visible (specimen PIN, no. 1758/86, 255, etc.), it is always located anterior to the old root. In Inostrancevia, the canine capsule is very wide, contains up to three capsules of replacement canines, which contain crowns of replacement canines at different developmental stages (Fig. 57a). 9. Ontogenetic Changes in Skull In theory, the ontogenetic study could have provided invaluable data on the phylogeny; however, only a very few taxa provide data on cranial changes in ontogeny. As the skull increases in size, all groups show an increase in ossification of the chondrocranium, some complication and reinforcement and sometimes fusion of sutures, thickening of membrane bones with the PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

943

(b)

(a)

(c)

Fig. 56. Tooth wear: (a) Suminia sp., specimen SGU, no. 104B/1351, right premaxillary, lingual view; (b, c) Suminia getmanovi Ivachnenko, 1994: (b) specimen PIN, no. 2212/99, left dentary, labial view; and (c) specimen PIN, no. 2212/31, skull fragment, right lateral view. Scale bar, 0.5 cm.

(a)

(b)

Pmx

dcs II

Mx

dbs II

dbs

dcs IV

fds

dcs III

dbs III

Fig. 57. Tooth replacement: (a) Inostrancevia latifrons Pravoslavlev, 1927, specimen PIN, no. 2005/1715, left maxillary, ventral view; and (b) Suminia getmanovi Ivachnenko, 1994, specimen PIN, no. 2212/18, left premaxillary and partial maxillary, lingual view. Designations: (dbs) replacement cheek tooth, (dcs) replacement canine, (fdc) fossa dentia canini superioris, (Mx) maxillary, (Pmx) premaxillary, and (II, III) tooth generations. Scale bars: (a) 0.5 cm and (b) 1 cm.

intensification of surface sculpture, and reduction of the relative orbital diameter. In Biarmosuchus tener, the development was traced from 16.5-cm-long skull to approximately 65–70-cm-long skull (Ivakhnenko, 1999). PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

In addition to the above mentioned features, the major character is a disproportionate growth of canines and, hence, an increase in the relative height of the canine capsule, particularly that of the upper jaw, which is con-

944

IVAKHNENKO (a)

(b)

(c)

Fig. 58. Ontogenetic changes in the skull of Biarmosuchus tener Tchudinov, 1960: (a) reconstruction based on holotype PIN, no. 1758/2; (b) reconstruction based on specimens PIN, nos. 1758/8 and 86; and (c) specimen PIN, no. 1758/1 (holotype Eotitanosuchus olsoni Tchudinov, 1960). Scale bar, 5 cm. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

945

(a)

(b)

Fig. 59. Ontogenetic changes in the skull of Titanophoneus potens Efremov, 1938: (a) lectotype PIN, no. 157/1; and (b) specimen PIN, no. 157/3 (holotype of Doliosauriscus yanschinovi Orlov, 1958). Scale bar, 5 cm.

nected with an increase in the height of the preorbital part of the skull. As the skull becomes twice as long, the canine length and the height of the facial plate of the PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

maxillary increase by somewhat more than 2.5 times, whereas the orbital diameter increases by a little more than 1.5 times. At the same time, both preorbital and

946

IVAKHNENKO

(a)

(b)

(c)

(d)

Fig. 60. Ontogenetic changes in the skull of Estemmenosuchus uralensis Tchudinov, 1960: (a) specimen PIN, no. 1758/300 (holotype of Zopherosuchus luceus Tchudinov, 1983); (b) reconstruction based on specimen PIN, no. 1758/79 (holotype Anoplosuchus tenuirostris Tchudinov, 1968); (c) reconstruction based on specimen PIN, no. 1758/331; and (d) specimen PIN, no. 1758/22. Scale bar, 5 cm. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

temporal skull regions increase in length by slightly more than twice due to the relative decrease in orbital diameter (Fig. 58). Titanophoneus potens is represented by two skulls, showing a moderate difference in size (42 and 55 cm long). However, as the skull grew, the relatively small pachyostotic expansions in the area of the parietal tubercle and above the orbits became a huge continuous pachyostotic expansion that occupied the entire frontoparietal region (Fig. 59). In Estemmenosuchus uralensis, the skulls ranging from approximately 10–12 to 60–65 cm of length were examined (Ivakhnenko, 2000a). Unfortunately, young individuals are only represented by fragments of the relatively massive frontoparietal region. Therefore, in addition to the common features listed above, it is only possible to indicate intense development during ontogeny of the postorbital frontoparietal pachyostoses (Fig. 60), which change from small thickenings to relatively high tubercles. In some very large individuals, tubercles become hornlike outgrowths with a lateral projection. The presence of skulls of approximately the same size, but having either tubercles or branching horns suggests that the latter belong to adult males, and the excessive development of horns is connected with sexual selection (Ivakhnenko, 2000a). The same developmental trend from small tubercles to laterally elongated projections was characteristic of pachyostotic thickenings in the zygomatic region. Note that the postorbital arch abruptly expands in ontogeny due to an increase in the orbital plate of the postorbital. During ontogeny, this part of the postorbital developed from a narrow postorbital crest to a wide plate, which covers externally the anterior part of the adductor fossa. It is interesting that the same process of expansion of the postorbital arch during ontogeny is also observed in the Phthinosuchidae, judging from specimens of different individual ages (Dinosaurus murchisoni, specimens PIN, nos. 1954/3, 296/1, and Viatkogorgon ivakhnenkoi, specimens PIN, nos. 2212/61, 4678/5). In addition, in this group, the zygomatic arch considerably expands in ontogeny mostly because of the ventral growth of the thickened jugal part. In almost all of these cases, age changes manifested in the general shape of the skull were originally regarded as taxonomic differences, and individuals distinguished by the size were described under different generic names, or even assigned to different families. This shows the necessity to analyze with caution external characters that may change during ontogeny. This is particularly true of specimens of different sizes coming from the same locality or stratigraphically identical localities. CHAPTER 4. CRANIAL MORPHOGENESIS In my opinion, the general skull design is determined to a great extent by the optimum relationships PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

947

(for a particular form) of inherited structures (the bone set and patterns of their articulation), necessary fenestration, which is connected with the presence of exits for sensors of the organs of sense, and fenestration or structures connected with the mechanical stresses produced by work of the jaw and cervical muscles. It is possible to regard all subsequent changes in the evolution of the uniform group or a number of sister lineages as an example of plastic deformations of interrelated structures. The initial skull archetype of Dinomorpha was undoubtedly formed at the level of Sphenacomorpha. This was the group that gave rise to the primarily high and relatively narrow skull, with distinct prevalence of the height over the width. This skull shape was probably connected with an early and principal constructional feature of Eotherapsida, i.e., with the origin of the jaw muscles from the anterolateral surfaces of bones of the ear capsule (the prootic braincase type). This structural character is connected with the majority of characters of the skull archetype of Dinomorpha. In particular, this concerns the formation of the united structure of the periotic, which combines the synoticum, opisthotica, and basioccipitale fused in a massive subvertical plate adjoining the prootic and basisphenoid. These features of the periotic have already been formed in Sphenacomorpha (see, for example, Romer and Price, 1940, text-fig. 10, and Modesto, 1995, textfig. 15, Edaphosaurus). The medial displacement of jaw muscles towards the braincase resulted in a significant decrease in size of the cavum epiptericum and, hence, enveloping of the lower part of the epipterygoid by muscles and a significant reduction of the epipterygoid. The same process of very primitive improvement of the jaw muscles (elongation) results in lowering the regions of the quadratum–articulare joint, with a distinctive curvature of the surangular. In the temporal region, the same process causes a considerable development of the caudal process of the postorbital, which reaches the squamosal and is located in the area of the greatest development of muscles. As a result, the zone of the former spiracular fissure (in this case, the fissura parapsida), which is weakened in tetrapods, is reinforced. Naturally, the bones of the lateral frame of the parietal shield (supratemporals) are in a very disadvantageous position from the functional point of view. In Sphenacomorpha, these bones form narrow bands coming onto the dorsal surface of the roof; however, in Dinomorpha examined in this respect, they are only retained in the shape of thin plates, the basic function of which is separation of the lateral surface of the paroccipital process of the squamosal by the apophysis paroccipitalis. In the majority of Dinomorpha, the supratemporals are probably completely reduced. Primitive Dinomorpha undoubtedly inherited the external crest of the angular bone (lamina reflexa); this suggests an important role of the pterygomandibular muscle, which in this design could not pass around the postdentary part of the lower jaw because of the presence of

948

IVAKHNENKO (a) (d)

(c) (b)

Fig. 61. Skulls of primitive Dinomorpha: (a) Nikkasaurus tatarinovi Ivachnenko, 2000, holotype PIN, no. 162/33; (b) Niaftasuchus zekkeli Ivachnenko, 1990, reconstruction based on holotype PIN, no. 3717/36, specimen PIN, no. 4543/20; (c) Microurania minima Ivachnenko, 1995, reconstruction based on holotype PIN, no. 4337/1; and (d) Ustia atra Ivachnenko, 2003, reconstruction based on holotype PIN, no. 4548/155 (skull outline, tentative reconstruction). Scale bar, 1 cm.

the periangular cavity on the external surface of the angular. In Sphenacomorpha, this cavity is represented by a small flat depression, which opens morphologically posteriorly and is slightly overlapped by the ala angularis, i.e., represents the cavum infradentalis. Representatives of Sphenacomorpha already have a high plate of the quadrate, which is closely connection with the quadratojugal; however, they lack the major character of the general design of Dinomorpha, i.e., a united QQJ-complex lacking a connection by the lateral suture with the squamosal (compare Romer and Price, 1940, pl. 7). It is evident that the best investigated Sphenacomorpha (Sphenacodontia, Edaphosauria) may not be taken for a group ancestral to Dinomorpha, although they are rather similar in appearance. As indicated above, these groups have the same general archetype, with primitive features combined with characters of high specialization. It is probable that the primitive Early Permian “Haptodontidae” display a design closely similar to the initial state of Dinomorpha; however, this group is currently represented by a very poorly preserved and poorly understood material. Even the validity of this taxon or its composition remain uncertain (see Reisz, 1986, p. 74). I examined the material of Haptodus longicaudatus (Credner, 1888) (= Palaeohatteria longi-

caudata, probably young individuals) in Staatlisches Museum für Mineralogie und Geologie (Dresden). The assignment to Sphenacomorpha is based on the following indirect arguments (mostly of specimen no. 306): the quadrate–articular region is displaced far posteriorly from the temporal fenestra; the ala angularis is very small relative to the lamina reflexa; the high symphyseal region of the dentary is positioned far anterior to the upper canine-shaped tooth, this suggests that the lower incisors were probably well developed. Unfortunately, in the material examined, it is impossible to check the basic character, i.e., the presence or absence of contact between the squamosal and quadratojugal. Additional material of the genus Haptodus, which is known from publications (H. bayleri Gaudry, 1886; H. grandis Paton, 1974; and H. garnettensis Currie, 1977), is either insufficiently preserved or described incompletely to establish true relationships between this group and Dinomorpha. However, the close affinity of this group to primitive Dinomorpha is indubitable. As follows from morphological analysis, the family Nikkasauridae should be regarded as the most primitive group of Dinomorpha (Fig. 61a). These small-sized animals probably display the structural archetype of the entire group, since they almost lack pronounced feaPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

tures of narrow specialization. An important specialized feature is only the presence of the upper canineshaped tooth in Reiszia (without an increase in the size of the lower incisors); however, similar structure of this region is known in some Sphenacomorpha (Haptodus longicaudatus; Secodontosaurus: Reisz, 1986, textfig. 16). The Nikkasauridae retain many primitive features, which undoubtedly show close affinity to Sphenacomorpha. For example, they have a large incisura interpterygoidea and paired vomer with rows of small teeth along the medial suture. The lower margins of the vomer are almost horizontal, the nasopharyngeal meatus is not formed as a separate structure. The lower jaw retains the praecoronoideum (coronoideum anterior). The processus basicranialis is very short and the basipterygoid articulation looks even more primitive than is typical for Sphenacomorpha (with long massive basipterygoid processes, which have a wide circular anterior parachondral surface). This region probably retained relative mobility. The braincase structure is also very primitive, with weakly fused bones of the periotic, a wide synotic, and a wide floccular fovea (plate braincase). At the same time, in the Nikkasauridae, the quadratojugal and squamosal do not form sutural contact; in the type specimen of Nikkasaurus tatarinovi (PIN, no. 162/33), the anterior margin of the crista superior of the squamosal remains smooth, without traces of articulation; the quadrate and quadratojugal were isolated and displaced anteriorly. The upper margin of the quadrate is thickened and forms the capitulum quadrati with a parachondral surface; this strongly suggests that the QQJ-complex was mobile. In addition, the quadrate– articular region has a standard mobility, although the surfaces of the lateral condyle and cavum infraarticularis are distinctly flattened in comparison with the surfaces of the medial condyle and cavum articularis. It is noteworthy that, in representatives of this group, the facies symphyseos of the lower jaw rami is longitudinally extended and has a smooth surface; this suggests some mobility between jaw rami. The jaws were probably capable of relatively complex movements. Animals had a bicrural–quadrate system of sound conduction; the stapes rests on the lamina tenticulata of the quadrate, a large and thin plate of the ala angularis is well developed. Regarding the relationship of the skull design of Nikkasauridae and other Dinomorpha, primary attention should be paid to the structure of the temporal region. This group is characterized by a high position of the base of the zygomatic process of the squamosal; therefore, the subapsid incisure is much more developed than in Gorgonopia and Dinocephalia, the zygomatic arch of which is horizontal in a primitive design and is often considerably lowered, covering the subapsid incisure. On the contrary, in Anomodontia, the incisure is strongly developed and its anterior margin spreads under the jugal. The dentition of Nikkasauridae is also primitive. They have many (more than 20) jaw PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

949

teeth, the anterior teeth are primitive, conical (of the centrocone type); only in the posterior third of the tooth row, they acquire additional cusps, changing into a primitive lateroconule type; however, they retain the shape of a piercing tooth rather than change into a primitive cutting long-leaf variant. Small-sized Microuraniidae probably had an unusual specialization developed based on a very primitive design (Fig. 61c). At present, they are represented by a very poor material, and their taxonomic position remains uncertain. Judging from available skull fragment of Microurania minima, the zygomatic process of the jugal is almost horizontal, with a slightly elevated upper margin. Hence, the zygomatic arch is approximately at the same level as in Niaftasuchidae or Ictidorhinidae and more lowered than in Nikkasauridae. In the original description, this was regarded as evidence of affinity of Microurania and primitive Rhopalodontidae (Ivakhnenko, 1995b); however, this contrasts with the absence of true alternative canines. Subsequently, this taxon was placed closer to Niaftasuchidae (Ivakhnenko, 2003c, p. 372), because the lateral margin of the postfrontal, which forms a suture with the postorbital, is somewhat thickened; this thickening was taken for crista postorbitalis. However, this suture is thickened in many Dinomorpha (apophysis postorbitalis of the postfrontal, a trace of unrealized parapsid fissure). Until new data are available, Microuraniidae should be regarded as a very primitive taxon closely related to Nikkasauridae. The dentition of Microuraniidae is somewhat unusual. The crowns of the upper anterior incisors expand strongly longitudinally, with ovate occlusal surfaces equipped with thin longitudinal ridges. The cheek teeth have a weak cinguloid, the lateroconules on the lateral borders are very small. The canine-shaped teeth are present in both upper and lower jaws, they correspond to a cheek tooth with a considerably increased apex (centrocone). The surface sculpture of Microurania is also unusual. Relatively smooth parietals and frontals (only background sculpture is observed) are combined with large, circular, closely located pits not only on the facial plates of the maxillaries but also on the entire surface of the nasals, extending onto the septomaxillaries. Thus, the skull of Nikkasauridae combines general primitive features of Eotherapsida and the characters undoubtedly corresponding to Dinomorpha; however, the latter are rather primitive, presumably corresponding to the initial state of the other lineages (Gorgodontia and Anomodontia). In fact, these groups differ primarily in the extent to which they use the subapsid incisure in the force skull design. In Gorgodontia, the zygomatic arch is at most horizontal, but is usually lowered, so that the formal presence of the subapsid incisure is of no significance for the development of jaw muscles (monofenestral design: Ivakhnenko, 2005b; this term is introduced to replace the obviously lame term

950

IVAKHNENKO

monoapsida: Ivakhnenko, 2003c); thus, only the increased synapsid fenestra functions as the temporal fenestra. On the contrary, Anomodontia use this incisure; it is sharply increased and comes under the zygomatic process of the jugal (difenestral design: Ivakhnenko, 2005b); thus, the functions of the temporal fenestrae are performed by the upper (synapsid) and lower (subapsid incisure) fenestrae. It is evident that the two designs are derived from the same initial archetype, in which the zygomatic arch is not yet lowered, does not cover the subapsid incisure, but is not positioned too high to provide the spread of the incisure under the jugal. This is precisely the state which is observed in Nikkasauridae and probably corresponds in morphology to the initial archetype of Dinomorpha; thus, based on formal reasons, Nikkasauridae should not be included in Gorgodontia or Anomodontia. The monofenestral skull design of Gorgodontia gave rise to at least two evolutionary trends, which are probably connected with the improvement of jaw muscles. The taxa of Gorgodontia, namely, Dinocephalia and Gorgonopia, correspond to different primary trends in the improvement of the jaw apparatus, in particular, the reinforcement of the lower jaw adductors. In Dinocephalia, the primitive skull design, with a vertically extended temporal fenestra is retained. The adductors were strengthened mostly through an increase in relative skull height. This involved the development of the anterior muscular portions, which passed into the postorbital region, displacing anteriorly the crest on the posterior margin of the postorbital (crista postorbitalis). This region has an anterodorsal depression bordered anteriorly by the crista postorbitalis, which moved anteriorly and medially in the course of evolution, approaching the extreme position on the frontal. The occipital plate of the squamosal remains almost vertical. This results in the formation of a vertical muscular block of posteriorly curved anterior portions of the adductors (anterosynapsid design: Ivakhnenko, 2005b). In Gorgonopia, the jaw muscles followed a different developmental trend, probably including the reinforcement of the posterior portions of the adductors. The occipital skull region is elongated by the posterior and posterolateral curvature of the occipital flange of the squamosal above the paroccipital process of the periotic. The obliquely positioned muscular block of the anteriorly curved posterior portions of the adductors was formed (posterosynapsid design: Ivakhnenko, 2005b). The crista postorbitalis is located on the caudal margin of the postorbital and sometimes even passes onto its medial surface; and in the evolution, the postorbital arch tends to expand and overlap externally the anterior margin of the temporal fossa. Certainly, Dinocephalia and Gorgonopia had the same the initial force skull design. However, judging from the most primitive known representatives of these evolutionary lineages and from the general tendencies in further development, the initial division into Dinocephalia and

Gorgonopia could have been connected with the formation of canineless taxa and taxa developing canines. Unfortunately, as usually occurs, primitive groups remain poorly known. Apparently, it is possible to regard Niaftasuchidae as a group most similar in morphology to the initial state of Dinocephalia, while Ictidorhinidae is most similar to that of Gorgonopia. Both groups have relatively weak, but distinct structural characters typical of the respective evolutionary trends. The Niaftasuchidae are the most primitive known Dinocephalia, represented by the only genus Niaftasuchus, the material of which is very poor (Fig. 61b). In this genus, the supraorbital fovea still occupies a small area on the dorsal surface of the postorbital, and the linea temporalis anterior is displaced slightly anteriorly. Primitive structural features probably include the preservation of the interpterygoid incisure and almost horizontal lower margin of the vomer; hence, the nasopharyngeal meatus is very weakly developed. Some dental features of Niaftasuchus are particularly important for our analysis. This genus has three strongly increased anterior incisors in the premaxilla and dentary. The incisors are strongly inclined anteriorly, have almost cylindrical crowns with flattened lingual surfaces. On these surfaces, longitudinal varicoid ridges are reinforced, becoming strengthening crests. The crown base of the incisor is also strengthened, thickened, and acquires a very weak cinguloid. The apices of incisors are dolabriform, sharpened by rubbing against antagonistic incisors. The cheek teeth have longitudinally extended crowns (short-leaf), with a few large lateroconules, form lateral borders of the mouth cavity. This taxon is characterized by an increased upper tooth that resembles in structure a longitudinally compressed incisor. The lower jaw curves slightly downwards, suggesting a relatively weak klinorhiny of the skull. The facial plate of the maxillary has widely spaced pits on the background of a relatively smooth surface, with infrequent foramina for blood vessels and branching grooves extending from these foramina. A primitive streptostyly of the QQJ-complex was probably retained. Although available data are rather poor, it is possible to propose that the skull structure of Niaftasuchus is similar to the initial design of East European Dinocephalia as a whole. The group represented in eastern Europe by Ulemosaurus (Fig. 62a) is close in morphology to Niaftasuchidae. This genus is distinguished by the large size and, hence, by the force skull design. Primarily, its skull is relatively narrow and high. The temporal fenestra remains extended ovate, the zygomatic arch is strongly widened, covering laterally the entire lower part of the temporal fossa. The necessity of retaining the relative braincase volume caused a considerable shortening of the paroccipital process. To compensate some narrowing of the temporal fossa, the jaw muscles expanded considerably into the postorbital region of the skull, with the development of the supraorbital fovea limited anteriorly by the crista postorbitalis of the postorbital. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

951

(c) (a)

(b)

Fig. 62. Skulls of Dinocephalia: (a) Ulemosaurus svijagensis Riabinin, 1938, lectotype PIN, no. 2207/2; (b) Deuterosaurus jubilaei (Nopcsa, 1928), reconstruction based on specimen PIN, no. 2629/1 (lower jaw based on D. biarmicus Eichwald, 1846, holotype PIN, no. 1954/1); (c) Titanophoneus potens Efremov, 1938, based on lectotype PIN, no. 157/1 and specimen PIN, no. 157/3. Scale bar, 5 cm.

The area for the muscle closely approaches the frontal. The streptostyly of the QQJ-complex completely disappears. The articular area on the squamosal for the complex levels off and expands, a gap between this area and the caudal surface of the complex disappears, and the quadratojugal, which is located in the same plane as the quadrate, tightly adjoins the flat lateral part of the expanded articular area. The QQJ-complex is bordered anteriorly by the expanded quadrate process of the pterygoid, which forms tight angular contact with the quadrate. Certainly, the absence of streptostyly has a significant effect on the structure of the quadrate–articular region. The condyles of the quadrate (condylus lateralis and condylus medialis) have convex cylindrical surfaces, and the mandibular cavum articularis and cavum infraarticularis are deepened and look like PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

obliquely longitudinal grooves with a depressed surface. The most significant reorganization in the palatal region is connected with the elevation of the choanal area, the formation of a very high nasopharyngeal meatus, and an increase in the bone massiveness. The pterygoid flanges are strongly thickened, but the palatal teeth on them and on the palatine and pterygoid tubercles are reduced, only two or three small, irregularly positioned teeth are retained. The klinorhiny is well developed, i.e., the anterior part of the palate posterior to the line of the pterygoid flanges is lowered and, hence, the prootic region of the periotic is lowered and elongated (klinorhinal braincase). This is accompanied by a ventral curvature of the anterior part of the lower jaw in the region of contact between the dentary and postdentary parts; in the sym-

952

IVAKHNENKO

physis the dentaries and splenials are fused. The cheek teeth become somewhat inclined in alveoli. The greatest changes of the jaw teeth occur in the incisive region. Their crowns are strongly widened, the centrocone is displaced labially, and a very wide cinguloid is formed on the lingual surface. In the crown of the replacement incisor, the cinguloid is bordered by small lateroconules; however, as the crown erupted from the alveolus, it was at once heavily worn by the antagonistic incisor. As a result, the occlusal surface of the upper incisor becomes somewhat compressed, while that of the lower incisor is convex; this strongly suggests the absence of horizontal movements of the jaw. Ulemosaurus has a canine-shaped tooth in the upper jaw. It is closely similar in morphology to the increased upper tooth of Niaftasuchus, which becomes more cylindrical, but retains longitudinal crests on the lingual surface. The animal has many relatively small cheek teeth, with longitudinally expanded crowns (short-leaf), without traces of natural wear by occlusion; they form a fence structure. This genus is characterized by a high cupolashaped pachyostosis in the frontoparietal region, with an obliterated sagittal suture and strengthened medial line of the skull. It is possible to regard the skull of Deuterosaurus (Fig. 62b) as a result of further transformation of the design described. In this taxon, the skull becomes relatively narrower and higher, with an even more significant decrease in length of the paroccipital process. The temporal foramen expands considerably. The parietal region of the skull roof is narrowed considerably anterior to the parietal foramen, such that the surface of the parietals free from muscles is limited to the margins of the tubercle surrounding the parietal foramen, and the supraorbital fovea for some portions of jaw muscles comes onto the frontal. The nasopharyngeal meatus is very high and probably separated from the mouth cavity by a soft palate covered with small teeth. However, the dental structure suggests independent evolution of this lineage beginning from a relatively low morphological level, probably shared with Ulemosaurus. The incisors of Deuterosaurus are identical in structural plan, but look much more primitive and less widened than in Ulemosaurus; it also has a lower canine-shaped tooth, and the upper tooth is elongated and shaped like a true predatory canine, although it retains a characteristic thickening at the transition to the collum, in place of the cinguloid. The cheek teeth are probably derivatives of the short-leaf dental type, thickened and specialized for crushing. Deuterosaurus is characterized by the development of a relatively narrow, high crestshaped pachyostosis mostly in the nasofrontal region. The skull of Titanophoneus (Fig. 62c) is undoubtedly an independent derivative of the same initial design, which changed in a similar, but somewhat different direction. As compared to the previous genus, the skull of Titanophoneus is elongated in the preorbital part, and shows a weak, probably inherited klinorhiny. The streptostyly is impossible; the posterior margin of

the quadrate process is widened, and forms tight angular contact with the quadrate. Hence, the lateral and medial condyles have convex cylindrical surfaces, and the cavum articularis and cavum infraarticularis are deepened and become oblique longitudinal grooves with a compressed surface. The connection of the lower jaw rami in the symphyseal region is strengthened by the fusion of the splenials and dentaries. The jaw muscles are distinctly strengthened. The caudal process of the postorbital is considerably elevated, raising the posterodorsal border of the temporal fenestra and, hence, significantly increasing the volume of the temporal cavity. In addition, the zygomatic arch deviates somewhat laterally, such that a small external portion of the adductor comes onto the external surface of the dentary and is bordered from below by the angular crest. The incisors become high pointed conical, but retain a weak cinguloid, and, as an abnormal variant, the incisor resembles in shape that of Ulemosaurus. The upper “canine” becomes flattened, and a thickening in place of the cinguloid is almost absent. At the same time, the lower “canine” is still similar to that of Deuterosaurus, although it is already somewhat flattened, but retains a finely serrated cinguloid. The canine-shaped teeth are permanent, and capsules for replacement canines are absent. The cheek teeth have a well-developed high centrocone, with a cutting border, but also retain a narrow cinguloid. Titanophoneus has a pachyostosis in the frontoparietal region, which in the shape of a flat shield covers almost all skull roof. In the three taxa examined, the surface of the facial plate of the maxillaries is almost smooth, with a greater or lesser number of relatively large foramina for blood vessels, which give rise to grooves extending on the bone surface. No trace of pitted sculpture is recorded. These genera have a distinct monocrural– postquadrate auditory region, with a large ala angularis, the lateral surface of which is smooth. A distinct evolutionary trend is represented by Microsyodon, Archaeosyodon, and Syodon (Figs. 63a, 63b). These taxa have a very unusual sculpture on the facial lamina. The bone surface is covered with small tubercles and ridges with a fine-porous surface. This surface structure is probably connected with the development of peculiar tactile sensory fields in the cheek region; however, this question remains poorly understood. Like the taxa considered above, Archaeosyodon and Syodon have a well-pronounced klinorhinal braincase and a well-developed supraorbital fovea for jaw muscles, so that the postorbital crest of the postorbital extends anteriorly up to the postfrontal; in Syodon, it reaches medially the parietal foramen. The temporal fenestra is widened considerably posteriorly. In Archaeosyodon and Syodon, the skull is nonstreptostylic, the posterior margin of the quadrate process of the pterygoid is widened and forms tight angular contact with the quadrate. Therefore, the surfaces of the jaw condyles of the quadrate are convex and cylindrical; those of the articular bone are deeply concave. The cheek teeth of Microsyodon are piercing, approach the PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

953

(a)

(b)

Fig. 63. Skulls of Dinocephalia: (a) Archaeosyodon praeventor Tchudinov, 1960, reconstruction based on specimens PIN, nos. 1758/95, 293, and 297; and (b) Syodon efremovi (Orlov, 1940), holotype PIN, no. 157/2. Scale bar, 5 cm.

long-leaf type, with a strengthened, almost cutting serrated posterior border; in Archaeosyodon, they are more circular in cross section and short; and in Syodon, they are extremely short, crushing. An attempt undertaken by the last genus to use cheek teeth not only as an element of a primitive grasping apparatus but also for a crushing treatment of food objects in the condition of almost undeveloped alternative occlusion resulted in certain disturbance of their position and an increase in the number of posterior teeth in the row, with the development of a relatively wide crushing field. The resultant increase in the pressure on the jaws of Syodon resulted in the fusion between the splenials and between the dentaries of the opposite jaw rami. The nasopharyngeal meatus is relatively low, and no trace of a soft palate is observed. The evolution of the upper canine-shaped tooth in this lineage is of special interest. In Microsyodon, the maxillary retains a precanine, and the canine-shaped tooth is flattened in cross section, has poorly pronounced edges, resembles in morphology the lower “canine” of Titanophoneus. The tooth is in an almost vertical position in the jaw, but considerably curved posteriorly in the distal part. In Archaeosyodon, the tooth is almost circular in cross section, with a smooth surface and a very weak smooth posterior longitudinal crest (border). It occupies an inclined position in the jaw (about 70°). In Syodon, the “canine” is cylindrical in cross section, lacks a trace of borders, sharply curves posteriorly, and occupies an inclined position in the jaw (about 60°). In all cases, replacement “canines” are absent, since no trace of capsules for replacement teeth is observed. The lower “canine” (if known) is relPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

atively small, with a thickening at the base in place of the cinguloid. Some cranial features are important for further consideration. The braincase of Syodon has a very welldeveloped fovea floccularis. This fossa is very deep, irregularly subrectangular, and considerably exceeds in relative volume the same structures not only in Dinocephalia but also in all Dinomorpha. On the contrary, the periangular cavity is very small in area and volume; the ala angularis is only slightly lowered relative to the lower margin of the jaw. The pachyostosis on the skull roof is not characteristic of this group, it is only limited to narrow ridges above the orbits, on the anterior margin of the parietal foramen, and along the lower margin of the zygomatic arch. The vascular system in the region of the frontal is only represented by one or two small foramina, with very weak imprints of branching vessels. Another unusual trend in the evolution of Dinocephalia is represented by Eotitanosuchidae and Alrausuchidae (Figs. 64a, 64b). In both groups, the temporal region shows a structure typical for Dinocephalia, with a relatively small, but well developed fovea supraorbitalis, restricted anteriorly by a high crest (crista postorbitalis ossis postorbitalis). The occipital plate of the squamosal is vertical, and the temporal fenestra is almost vertical, narrow ovate, in Biarmosuchus, with a slightly elevated posterodorsal margin. Despite a primitive structure of this region, the klinorhiny of the braincase is already distinct, with an abruptly lowered basisphenoid and a large angle between the planum pituitaria and clivus; this strongly suggests the origin from phytopha-

954

IVAKHNENKO

(b)

(a)

Fig. 64. Skulls of Dinocephalia: (a) Alrausuchus tagax (Ivachnenko, 1990), reconstruction based on holotype PIN, no. 3706/10 and specimens PIN, nos. 4659/8 and 3586/14; and (b) Biarmosuchus tener Tchudinov, 1960, reconstruction based on specimens PIN, nos. 1758/1, 86, and 307. Scale bar, 5 cm.

gous ancestors. The preorbital part of the skull is elongated. The skull almost lacks pachyostotic expansions, except for a thickening on the lower margin of the zygomatic arch. The primitive state of the group is evident from certain general structural features. An elevation of the lower margin of the vomer with the formation of the nasopharyngeal meatus are almost absent. A soft palate was undoubtedly absent, judging from the skin teeth described above on the lateral surface of the vomer of Biarmosuchus. The lower jaw retains the praecoronoideum (coronoideum anterior). The incisors have very weak cinguloids and are similar in structure to the incisors of Niaftasuchus. The cheek teeth are also very primitive, of the long-leaf type, only with a somewhat strengthened posterior crest, lack a trace of cinguloids. The main distinctive feature of the group is connected with the dentition, the development of alternative canines, i.e., the presence of both upper and lower canines which form an integrated system for food treatment. The canines are identical in shape and similar in size, strongly flattened in cross section, have welldeveloped anterior and posterior borders with cutting serration, and lack a trace of cinguloid or thickening in its place. Additional longitudinal edges are also absent. Certainly, these canines could have developed from

primitive cheek teeth through the growth of the elongated “leaf” and multiplication of bordering lateroconules. In Biarmosuchus, the canines are very long, saber-shaped, slightly curved, provided with a capsule for the replacement canine. The canines of Alrausuchus are much shorter relative to the skull size and curved much more abruptly. The development of alternative canines, which are very similar in structure to the canines of predatory Gorgonopia, emphasizes the primitive state of this group, because they could have developed only before the formation of cinguloids on the cheek teeth. The major distinction of Eotitanosuchidae from Alrausuchidae consists in the preservation and even better development of the initial streptostyly of the QQJ-complex. The capitulum quadrati is massive and widened, with a perichondral dorsal and partially caudal surfaces; it enters with a significant gap into a deep fossa quadratica on the squamosal. The processus quadrati of the pterygoid does not reach the quadrate bone; the same is true of the thin posterior margin of the basis epipterygoideus. The development of the streptostyly consists in interlocking the mobility in the jaw joint, and the preservation of articulation of the lower jaw only in the region of contact between the capitulum PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

quadrati and the squamosal. The wide flat surface of the cavum infraarticularis tightly adjoins from below the flat surface of wide medial condyle of the quadrate, excluding completely movements in the joint; the upper margin of the cavum articularis forms the posterior margin of the surangular, which curves somewhat posteriorly, forming the anterior tubercle of the surangular, which adjoins the anterior surface of the quadrate and contributes to the locking of its movements. Biarmosuchus displays a typical bicrural–quadrate design of the auditory apparatus. The distal end of the stapes is extended and adjoins the lamina tenticulata of the quadrate; the bone is in the shape of a thin and wide frame, with a very large intercrural foramen. The stapes is mobile, the perilymphatic duct and its paroccipital sinus are present. The ala angularis is thin, wavy, with a longitudinal ridge above the fissura interna; on the posterolateral margin of the squamosal, the crista posterior is well developed. All features described are also characteristic of Alrausuchus; however, it should be regarded as a taxon that has secondarily lost streptostyly of the QQJ-complex. In fact, the quadrate of the same shape is combined with a well ossified caudal part of the basis epipterygoideus, a massive plate of which rests against the bone, locking its mobility. The capitulum quadrati is distinct, but tightly adjoins the walls of the fossa quadratica. The restoration of mobility in the quadrate–articular joint is peculiar. The tubercle on the posterior margin of the surangular (tuberculum anterior) becomes low, and the cavum articularis is widened considerably. It contains the lateral condyle of the quadrate, while the flattened medial condyle is highly elevated and reaches a flat surface of the cavum infraarticularis only in the case of extreme posterior abduction of the lower jaw. Thus, only the cylindrical lateral condyle plays the role of a support of the lower jaw. The auditory region is also unusual. It is similar in detail to this region of Biarmosuchus, but the caudal surface of the quadrate acquires the sulcus stapedialis and it contains the lower part of the caput stapedis. Thus, the group regarded here as the taxon Dinocephalia is characterized by predominant development of the temporal muscles in the postorbital region. The origin of some muscles expands from the internal surface of the body of the postorbital onto the dorsal surface; the marginal crest (crista postorbitalis) extends anteriorly for a greater or lesser extent, reaching the frontal in extreme cases. As a result, the dorsal surface acquires a muscular area (supraorbital fovea) posterodorsal to the orbit. Apparently, this process was initiated by the necessity to increase the pressure in the incisive region of the jaws, since the most primitive known taxa, such as Niaftasuchus, have a strengthened incisive apparatus. The lineage most similar in morphology to the primitive design is represented in eastern Europe by Ulemosaurus, the incisive region of which shows further strengthening and specialization. A trend towards the use of the canine-shaped teeth is to a certain extent manifested in Deuterosaurus and even more proPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

955

nounced in Titanophoneus. In general, the dental apparatus of Deuterosaurus is derivable from an ancestor resembling Ulemosaurus; however, it had somewhat less specialized incisors, i.e., it evolved from taxa at a more primitive stage. The same concerns Titanophoneus, which is also primitive in incisor structure. Thus, the genera considered undoubtedly represent a single evolutionary trend, whereas the Archaeosyodon–Syodon group probably differentiated at an even more primitive level and its dentition followed a distinct direction of specialization, which displays extreme manifestation in Syodon. The trend represented by Eotitanosuchidae and Alrausuchidae is probably even more primitive in origin. In any case, the formation of true alternative canines, which are characteristic of these groups, could have developed only from the level of cheek teeth without cinguloids. In contrast to Dinocephalia, the most primitive known representatives of Gorgonopia have well-pronounced alternative canines. On the contrary, canineless taxa that could have been assigned to this group based on the structure of the temporal region are not known. This suggests that the early differentiation of teeth into incisors, canines, and postcanines was the character that determined the skull evolution in this group. Unfortunately, true primitive taxa of this group have not been recorded. They are similar in morphology to Ictidorhinidae; however, the latter is represented by late and specialized taxa. In eastern Europe, the family Ictidorhinidae is only represented by isolated dentaries; however, Sigogneau (1970a, 1970b) provided a relatively thorough description of South African Rubidgina Broom, 1942. The temporal region of this animal is primitive; the temporal fenestra is relatively narrow, vertical ovate; the postorbital arch is somewhat widened, with a marginal crista postorbitalis of the postorbital (Fig. 61d). The short zygomatic arch is only slightly lowered (Sigogneau, 1970a, text-fig. 1). The cheek teeth are primitive, of the long-leaf type, with many small lateroconules on the anterior and posterior borders. Longitudinal crests or edges reinforcing the crown are absent. It remains uncertain whether or not the taxa described in the literature retain streptostyly of the QQJ-complex; however, in East European taxa, such as Biarmosuchoides and Ustia, the symphyseal surfaces of the dentaries are longitudinally extended and have a smooth surface, like in Nikkasauridae; this is indirect evidence of the presence of streptostyly. In Rubidgina, the processus quadrati is figured as a very short process terminating short of the quadrate (see Sigogneau, 1970a, text-fig. 4); this bone has a wide lamina tenticulata, which is typical of streptostylic taxa. Lemurosaurus Broom, 1949, which was previously assigned to Ictidorhinidae, should be included in Burnetiidae based on new data (see Sidor and Welman, 2003). As the primary trend in the evolution of Gorgonopia was the formation of the jaw apparatus with alternative

956

IVAKHNENKO (a) (d)

(b)

(c)

Fig. 65. Skulls of Gorgonopia: (a) Dinosaurus murchisoni (Fischer, 1845), reconstruction based on specimen PIN, no. 1954/3 (young specimen); (b) Kamagorgon ulanovi Tatarinov, 1999, reconstruction based on holotype PIN, no. 4312/1; (c) Inostrancevia alexandri Amalitzky, 1922, lectotype PIN, no. 2005/1587; and (d) Suchogorgon golubevi Tatarinov, 2000 (on: Ivakhnenko, 2005b, text-fig. 6). Scale bar, 5 cm.

canines, the most primitive state is retained in Phthinosuchidae (Dinosaurus, Fig. 65a). The weak lowering of the zygomatic arch is masked in these animals by an intense downward growth of its ventral margin, which frequently has a pachyostotic expansion and covers dorsally the region of the recessus alae angularis. A similar expansion is observed on the postorbital arch, both expansions intensely developed in ontogeny. As the postorbital arch expanded, its posterior margin covered externally the anterior part of the temporal fossa; thus, the crista postorbitalis of the postorbital was displaced anteriorly on the internal (intracranial) surface. The posterodorsal border of the temporal fenestra is elevated; the dorsolateral margin (ala temporalis) of the squamosal, which borders posteroventrally the temporal fossa, is to a greater or lesser extent curved

externally (somewhat laterally and posteriorly), increasing the volume of the upper part of the cavity containing muscles. As a result, the plane of the temporal fenestra occupies a somewhat dorsolateral position (Fig. 65b). This is particularly well pronounced in Rubidgeidae (compare Dinogorgon: Sigogneau, 1970b, pl. 75, fig. b). In fact, the most significant differences between Phthinosuchidae and Rubidgeidae consist in the extent to which this design is developed. Other differences are connected with the structure of the palatal region (Phthinosuchidae have relatively wide palatal tubercles with many small teeth) and dentition (a decrease in the number of postcanines in Rubidgeidae). The postcanines are widened at the crown base (in the region of the collum dentis), somewhat curved posteriorly, reinforcing the posterior cutting border. The canines lack a conPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

striction in the region of the collum dentis and probably develop by an increase in size of a cutting cheek tooth of this structure. The Phthinosuchidae retain the plate braincase and acquire capsules for replacement canines. The facial surface of the maxillary has a pitted sculpture superimposed on the background sculpture. Available material suggests that the group had a streptostylic QQJ-complex; in any event, the quadrate process of the pterygoid does not reach the quadrate, and the capitulum quadrati is thickened and enters a large fossa quadratica of the squamosal (Viatkogorgon). The structure of the sound-conducting apparatus has not yet been examined; however, the quadrate of Viatkogorgon has a very wide, almost parasagittal lamina tenticulata and the group was probably characterized by the bicrural–quadrate design. The family Inostranceviidae is probably a result of a peculiar specialization of this lineage (Fig. 65c). The skull is elongated in both preorbital and temporal regions. The postorbital and zygomatic arches are moderately widened, to a much lesser extent than in Phthinosuchidae and particularly in Rubidgeidae. The elongation of the skull is accompanied by relative expansion of the temporal region; this increases the volume of the temporal fossae and, hence, the paroccipital process becomes very narrow and long. In Inostrancevia, the most significant changes are observed in the jaw apparatus. The upper and lower alternative canines are considerably elongated, almost equal in length, relatively weakly curved. The anterior and posterior borders are sharp, well developed and have rows of small cutting lateroconules of a uniform regular shape. The replacement of canines is evident, and a wide capsule of the upper canine sometimes has up to three additional rudiments. The incisors form a semicircular cutting line and, along with the alternative canines, participate in the united cutting jaw system. The number of upper postcanines is reduced (to six), the lower postcanines are completely reduced. The line of the upper postcanines is obliquely ascending, and the teeth are inclined posteriorly. The strengthened canines and incisors and strong jaw muscles cause the reinforcement of the sagittal suture anterior to the region of the frontoparietal suture; in this area, the suture becomes very sinuous. However, the maxillary is weakly attached to the neighboring bones (nasal, prefrontal, and lacrimal); this is particularly distinctly pronounced in the region of the incisivomaxillary suture, where a mobile joint rather than a sutural articulation is observed. An unusual attachment to the palatine through a few parallel crests is observed. This supports presumable functional significance of the limited mobility of the maxillary, which is probably connected with unusual work of the jaw apparatus. The QQJ-complex is streptostylic; the pterygoid and epipterygoid do not reach the quadrate; the capitulum quadrati is well developed and enters the fossa quadratica of the squamosal; the quadratum– articulare joint is immobile. The lower jaw rami are divided into weakly connected dentary and postdentary PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

957

parts. The functioning of this complex, unusually arranged apparatus necessitates a special study. It is probable that the combination of the streptostyly, weak connection within the lower jaw rami, and the arrangement of postcanines is attributable to swallowing large pieces of food. The nasopharyngeal meatus is very weakly developed, the palatines lack a trace of palatine teeth or tubercles. The external surface of the facial plate of the maxillary lacks a pitted sculpture, the bone has only relatively large foramina for blood vessels, with short grooves extending from them. It is noteworthy that the walls of the braincase lack a trace of fovea floccularis. The sound-conducting system was probably of the bicrural–quadrate type, although the structure of the stapes is not known. The ala angularis is thin and wavy, only slightly deviates from the body of the angular bone, with a wide, posteriorly open recessus alae and increased fissura externa. The retroarticular process is very short and massive. It is probable that Gorgonopidae (Pravoslavlevia, Sauroctonus, Suchogorgon: Fig. 65d) are also connected in origin with taxa resembling primitive Phthinosuchidae. The basic difference is the weak expansion of the zygomatic and postorbital arches. However, as mentioned above, these features developed in the ontogeny of Phthinosuchidae; therefore, it is possible to regard Gorgonopidae as a derivative of a somewhat more primitive group or usual Phthinosuchidae that underwent fetalization in these parameters. In Gorgonopidae, the dorsolateral margin (ala temporalis) of the squamosal, which limits posterodorsally the temporal fossa, curves strongly posterolaterally; this considerably increases the volume of the upper part of the cavity. At the same time, the temporal region is elongated considerably posteriorly; this is connected with a welldeveloped basicranial process. The group is characterized by a relatively large, rhombic preparietal (interparietal) in the middle part of the coronal suture. This bone undoubtedly reinforces the primarily weak sagittal suture by fastening and an increase in size of the ossa suturarum, which sometimes occur (although not necessarily present) in Rubidgeidae. The maxillary is very weakly attached to the neighboring bones, as in Inostrancevia, including the area of the sutura incisivomaxillaris and the palatine; this is probably connected with the kinesis of the maxillary (Ivakhnenko, 2005b, p. 439). The nasopharyngeal meatus above the region of the naria interna is high, although no traces of the soft palate are observed; the palatal tubercles are narrow, with a few small teeth. The streptostyly of the QQJ-complex is very well developed; the fossa quadratica is deep, with a widened upper margin, such that the capitulum quadrati enters the fossa quadratica with a considerable gap; and the quadrate process, which extends to the quadrate, is very short and forms a pointed, lowered hamulus pterygoideus at the caudal end. Consequently, the quadratum– articulare joint is immobile, the lower surface of a wide medial condyle is completely flat and rests on a flat sur-

958

IVAKHNENKO

face of the cavum infraarticularis, and the surangular has a well-developed anterior tubercle, which rests against the anterior surface of the lateral condyle. The lower jaw rami are divided into dentary and postdentary parts, the first only overlaps the second. Among all the taxa investigated, only Gorgonopidae have a considerably developed coronoid apophysis of the dentary, which extends posteriorly and somewhat dorsally beyond the contact area with the surangular. An unusual feature of this group is a very short ascending process of the premaxilla (spina nasalis), which terminates short of the anterior margins of the nasal. The pitted sculpture on the facial surface of the maxillary is very well pronounced. The system of alternative canines is well-developed; however, as compared to the skull size, the canines are much smaller and more curved than in Inostrancevia; they and the semicircle of incisors form a united apparatus, which cut off large pieces from the body of prey. The postcanines are reduced in number, weakened, and inclined posteriorly; this is probably connected with swallowing large pieces of food, as in Inostrancevia. The capsules of replacement canines are always well developed. The braincase has large floccular foveae. The sound-conducting system of Gorgonopidae is considerably changed compared to the typical bicrural–quadrate design (see above). Possibly, the entire flat postdentary part of the lower jaw functioned in this system as a very large tympanum. It considerably increased in area due to soft tissue stretching between the posterior margin of the ala angularis and long, anteriorly curved retroarticular process. The upper part of a very thin ala angularis adjoins the surface of the angular bone and, hence, the recessus alae is almost absent. It is possible that the development of this unusual design was connected with the existence of functional division of the lower jaw rami into dentary and postdentary parts. In other respects, the auditory system is constructed according to the usual bicrural– quadrate scheme, with a distinctly mobile stapes and a narrow vertical distal area of the caput stapedis, which adjoins a very wide lamina tenticulata of the quadrate; this retained contact between these bones when the quadrate moved. The fenestra rotunda is very well developed between a semicircular incisure in the basis stapedis and the sinus paroccipitalis on the lower surface of the paroccipital process of the opisthotic. A particular evolutionary lineage is represented by Burnetiidae and Estemmenosuchidae, which are similar to each other. Primitive members of these families are known, but represented by very poor material; therefore, the origin and relationships of this group are only tentatively reconstructed. In the general pattern of the skull structure, the group is undoubtedly similar to primitive Phthinosuchidae (such as Dinosaurus). It is noteworthy that, in the Estemmenosuchidae, the postorbital and zygomatic arches expand in the course of ontogeny (in Burnetiidae, ontogenetic changes are not known). The basic distinctive feature consists in the fact that, in this evolutionary lineage, the streptostyly of

the QQJ-complex probably disappeared very early. As a result, these animals retained primitive, relatively low occiput and their temporal cavity increased in volume mostly by the growth in posterior direction. The absence of streptostyly provided gradual development of the klinorhiny. This causes some external similarity in skull structure of typical representatives (particularly Estemmenosuchus) to klinorhinal Dinocephalia. This has been long used as a character determining the taxonomic position of the group. The state of the general skull pattern is probably most similar to primitive Burnetiidae (Proburnetia, Fig. 66b). This is evident from a comparison with Hipposauridae, a closely related and undoubtedly more primitive group (see Sigogneau, 1970b, pl. 83). The development of klinorhiny (the basipalatal angle is about 20°), which is manifested in the lowering of the anterior part of the skull, is accompanied by the development of a sharp bend under the orbit, which is connected with the preservation of the orientation of jaw muscles. The temporal fenestra is relatively small, slightly expanding posteriorly. The postorbital arch is widened; in the Burnetiidae, the postorbital and zygomatic arches have pachyostotic expansions, usually with the formation of a more or less developed tubercle on the posteroventral border of the temporal fenestra. The canines retain a primitive flattened shape, with distinct cutting borders, but are poorly developed; therefore, the symphyseal region of the lower jaw and the entire dentary part are relatively low. Characteristic features of the Burnetiidae are the pachyostotic expansions on the skull roof, projections and thickenings on the anterior border and, particularly, the upper border of the orbit. The outgrowths above the upper border of the temporal fenestra are usually massive; they are probably located in the region of the posterolateral edges of the parietals (the sutures between these bones are completely obliterated). These and supraorbital outgrowths are twined by branching grooves of blood vessels. In addition, a longitudinal pachyostotic crest usually extends from the region of the lambdoid suture to the anterior margin of the nasal and reinforces the sagittal suture. The teeth are of the long-leaf type, adapted for piercing rather than cutting. The structure of the auditory system is probably very interesting; however, only poor data on this topic are presently available. Judging from the material examined, Proburnetia had a bicrural stapes, the postoccipital lamina on the lateral margin of the paroccipital process was almost undeveloped; however, the posterolateral margin of the squamosal lacks a crista posterior; therefore, the postquadrate region is open laterally and the fossa elliptica is well developed. This mixture of characters of the bicrural–quadrate and monocrural–postquadrate designs could have provided significant data on the functions of this structure at a transitional stage; however, available material is insufficient for detailed examination. It should be noted that the stapes could have been mobile, judging from the presence in Niuksenitia of a very large pocket of the PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

959

(a)

(b)

(c)

Fig. 66. Skulls of Gorgonopia: (a) Phthinosaurus borissiaki Efremov, 1940 holotype PIN, no. 164/7 (skull outline, tentative reconstruction); (b) Proburnetia vjatkensis Tatarinov, 1968, holotype PIN, no. 2416/1; and (c) Estemmenosuchus mirabilis Tchudinov, 1968, holotype PIN, no. 1758/6. Scale bar, 5 cm.

cavum perylimphaticum, which occupies more than one-third of the extent of the paroccipital process. The taxa described are undoubtedly closely related to Rhopalodontidae, which are only represented by lower jaw fragments. Judging from the jaw structure, the klinorhiny was better developed, the ventral curvature of the lower jaw is well pronounced (Phthinosaurus, Fig. 66a); in this connection, the cheek teeth are positioned obliquely in alveoli. The teeth are widened at the crown, displaying a typical short-leaf shape, and lack a trace of natural wear; they probably functioned as a fence structure. Judging from the poor available material, the basic difference from Burnetiidae and Ictidorhinidae is the strongly developed canines and, hence, the high dentaries, particularly in the symphyPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

seal region. However, the canines are circular in cross section and have distinct longitudinal edges. The marginal borders are somewhat better developed than the edges, but almost lack a cutting serration, which is represented by widely spaced, irregularly scattered thickenings along the border. The taxa described are similar in morphology to Parabradysaurus, which is represented by jaw fragments and isolated teeth. This is a much larger animal; its canine (or tusk) is circular in cross section, has only weak edges, the marginal borders are similar to the edges. The dentary is very massive, high, expanded, has a relatively wide longitudinally extended labial area; therefore, the postcanine row is displaced lingually and passes at the level of the internal edge of the alveolus of canine.

960

IVAKHNENKO

The same trends are even stronger manifested in Estemmenosuchus (Fig. 66c). These very large animals have almost circular in cross section canines–tusks, without a trace of cutting borders or edges. It is noteworthy that the incisors also lose cutting borders and acquire weak edges at the base. The tusks are very large, the dentaries are high, particularly in the symphyseal region. These bones are widened, with a wide labial area; therefore, the postcanine row is displaced far inside from the internal edge of the alveolus of canine and even passes anteriorly somewhat lingual to the canine. The cheek row has an increased number of teeth (about 25). The structures such as cinguloids are undeveloped; in some teeth, the lower lateroconules extend for a short distance onto the lingual surface. The cheek teeth lack a trace of natural wear or occlusion, form a typical fence structure. Estemmenosuchus displays a well-pronounced klinorhiny, the basipalatal angle between the planes of the braincase and palate is 40°. However, in contrast to the klinorhinal Dinocephalia considered above, the basic bend of the palate is approximately in line with the pterygoid flanges; therefore, the lowering at the level of the basipterygoid articulation is much less and the basisphenoid part of the periotic is not lowered (klinorhinal braincase is not formed). As in Dinocephalia, klinorhiny is connected with the formation of a distinct nasopharyngeal meatus, although it developed somewhat unusually. The plate of the vomer shows an only slight dorsal curvature, while the areas of the ala ventralis and eminentia interchoanalis are lowered and expanded, separating the nasopharyngeal meatus from the mouth cavity. This region of the vomer is particularly wide in Estemmenosuchus uralensis and even has small teeth on the margins. Estemmenosuchus lacks streptostyly; the widened posterior margin of the quadrate process rests on the quadrate and, hence, the quadratum–articulare joint has convex lateral and medial condyles and deepened cavum articularis and cavum infraarticularis. The temporal fenestra expands posteriorly and ventrally, considerably increasing the volume of the cavity for muscles. In Estemmenosuchus mirabilis, some muscular portions even come onto the external surface of the dentary. As a result, the zygomatic arch deviates somewhat from the lateral surface of the dentary, and a narrow smooth area for the muscle is restricted from below by a rugose crest on the upper margin of the angular bone. The pachyostotic expansions on the skull roof are very well developed, particularly in the supraorbital and parietal regions; tubercles are also present on the nasals and the facial plates of the maxillaries. The pachyostosis develops in the course of ontogeny, high hornlike outgrowths expand in the region of the posterolateral corners of the parietals and jugals. The surface of the outgrowths is twined by grooves for blood vessels; possibly, they primarily contributed to the thermoregulatory function. However, the fact that excessively developed outgrowths on the parietals and jugals

are absent from some large adults (presumable females) suggests that these structures could have played a role of sexual distinctions. The sound-conducting system of Estemmenosuchus is similar to that of Burnetiidae; the occipital surface of the squamosal also has a large fossa elliptica; the postquadrate cavity is open laterally; on the posterior surface of the quadrate, somewhat above the jaw condyle, there is a sulcus stapedialis, which contained a wide caput stapedis, losing distal contact with the quadrate. The stapes retains the bicrural structure. However, the postoccipital lamina develops, restricting dorsally the postquadrate cavity. This review shows that, in the Gorgonopia lineage, the jaw musculature was improved by strengthening the posterior portions of the adductors. This undoubtedly correlates with early development of differentiated dentition with alternative canines. After the evolutionary stage similar in morphology to Ictidorhinidae, two major trends were formed. In the first, including the Phthinosuchidae, Rubidgeidae, Inostranceviidae, and Gorgonopidae, the upper part of the temporal fenestra is widened, expanding posteriorly. This trend reaches the extreme state in the Phthinosuchidae–Rubidgeidae lineage, in which the zygomatic and postorbital arches expand, such that even a fenestra is formed that extends not only laterally, but also substantially dorsally. It is possible to regard the Inostranceviidae as a particular lineage differentiated from the level of Phthinosuchidae, with the skull elongated and widened in the occipital region and slightly widened zygomatic arches; as a result, the temporal fenestra is widely open both dorsally and laterally. Probably, the initial level of Gorgonopidae was somewhat more primitive, since they retained a primitive lateral orientation of the temporal fenestra, which only increased in relative size but already had a distinct dorsal and posterior lengthening. The second trend, comprising Burnetiidae and Estemmenosuchidae, retains a primitive structure of the temporal region, with the primarily relatively small laterally oriented temporal fenestra; the cavity for muscles increased in volume mostly by posterior and somewhat ventral expansion. The Burnetiidae lineage is considerably less specialized in dental structure than Estemmenosuchidae. The two lineages possibly evolved independently from primitive ancestors and followed different adaptive routes. Thus, the groups assigned to Dinocephalia and Gorgonopia display interesting examples of undoubtedly parallel developments. For example, Biarmosuchus, a very large animal that has retained and developed the klinorhiny and acquired increased canines, is surprisingly similar in a number of structural features to largesized Inostrancevia. In particular, they are similar in the structure of long, flattened, and slightly curved canines. At the same time, the smaller Alrausiuchus has shorter and more strongly cured canines, like Gorgonopidae of approximately the same size. Large Estemmenosuchus PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

961

(a)

(b)

(c)

Fig. 67. Skulls of Anomodontia: (a) Otsheria netzvetajevi Tchudinov, 1960, reconstruction based on holotype PIN, no. 1758/5 (lower jaw based on Venyukovia prima Amalitzky, 1922, lectotype PIN, no. 48/1); (b) Ulemica invisa (Efremov, 1938), reconstruction based on holotype PIN, no. 157/5, specimen PIN, no. 1116; and (c) Suminia getmanovi Ivachnenko, 1994, reconstruction based on holotype PIN, no. 2212/10 and specimen PIN, no. 62. Scale bar, 2 cm.

has lost streptostyly and acquired klinorhiny and strong pachyostosis of cranial bones. In these parameters, it has become similar to, for example, Ulemosaurus (and particularly to Deuterosaurus, with canine-shaped teeth), although these taxa differ in the general structural pattern (see above). Anomodontia represent a distinct evolutionary trend of Dinomorpha, with an increased subapsid incisure, which is partially bordered by the zygomatic process of the jugal. Therefore, the base of the zygomatic process is positioned much higher than the fossa quadratica; the zygomatic arch is elevated and the temporal fenestra is represented by two openings; the upper is the synapsid fenestra and the lower is the subapsid incisure (difenestral design: Ivakhnenko, 2005b). This design is not unique to Eotherapsida; in Edaphosaurus, the subapsid incisure is well-developed and extends far under the jugal (Modesto, 1995, text-fig. 1). Certainly, there is no reason to discuss the hypothesis of the direct relationship between these two groups; this is undoubtedly an interesting example of parallel development in remote lineages. Some features of Anomodontia are connected with the development of the subapsid incisure. As the incisure developed, some external portions of jaw muscles came onto the lateral surface of the dentary; therefore, a special extended area restricted from below by a crest is formed on the dental ramus. It is probable that PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

the well-developed incisure contrasts with streptostyly; in all taxa examined in this respect, the expanded posterior margin of the quadrate process of the pterygoid rests against the anterior surface of the QQJ-complex. As is observed in the other groups of Dinomorpha, the disappearance of streptostyly is usually accompanied by transition to a monocrural–postquadrate design of the sound-conducting apparatus. The same process was probably characteristic of Anomodontia. In all taxa of this group, the stapes of which was examined, the crus posterius is reduced. The higher groups acquire the fossa elliptica and lamina postoccipitalis, which restricts the postquadrate cavity from above. Of East European groups investigated, Venyukovia and Otsheria show the most primitive design. Unfortunately, these taxa are very poorly understood, as frequently observed in primitive taxa. It is possible that they are synonyms, since Venyukovia is only represented by a dentary, while Otsheria is represented by a very poorly preserved skull with heavily damaged tooth crowns and without a lower jaw. Judging from available material, the zygomatic arch of Otsheria is moderately elevated (Fig. 67a), at most to the level of the lower third of the orbit; the cavity for muscles is only slightly expanded posteriorly into the region above the quadrate. The incisive plate of the premaxilla is narrow, the bony choana occupies a position usual for tetrapods.

962

IVAKHNENKO

The nasopharyngeal meatus is well developed due to a highly elevated lower margin of the vomer. The most remarkable and significant feature is the elongated cylindrical incisors. They are inclined anteriorly, dolabriform, have sharp flattened apices, which were sharpened by antagonistic incisors and, as a result, also became sharpened. The serrated marginal borders were displaced on both lingual and lateral sides, forming cinguloid structures on both sides. The cheek teeth (8–10) are high and cylindrical, similar to the initial short-leaf type. However, they have well-developed wear facets, which were undoubtedly formed by rubbing against the teeth of the opposite jaw. On different teeth, these facets are located in somewhat different planes; this is evidence of only vertical movements of the jaw. In Otsheria, the vomer curves highly dorsally; therefore, a welldeveloped nasopharyngeal meatus is formed. It remains uncertain whether or not this group retained streptostyly. In the only known skull of Otsheria, the internal surface of the squamosal at the junction with the quadrate probably lacks a distinct fossa quadratica for the capitulum quadrati, and the plate of the quadrate process of the pterygoid is well developed; however, the QQJ-complex was isolated during maceration. The occipital surface of the squamosal lacks a fossa elliptica; the surface looks somewhat convex. The postoccipital lamina is almost absent; there is only a small thickening in the lower part of the distal surface of the paroccipital process. Unfortunately, neither stapes nor QQJ-complex have been recorded; however, the auditory apparatus was probably more primitive than the monocrural–postquadrate type. In Ulemica (Fig. 67b), the zygomatic arch is only slightly elevated, even weaker than in Otsheria, but the posterior margin of the subapsid incisure extends into the region above the quadrate, and the part of the squamosal between the base of the zygomatic process and fossa quadratica is in the shape of an ovate depression open anteriorly and laterally; this increases considerably the volume of the lower part of the temporal fossa. The palatine plate of the premaxilla expands, displacing somewhat the anterior margin of the choana ossica to the line of the anterior edge of the maxillary. The incisive part of the dental apparatus is similar in structure to that of the previous group; however, the cheek teeth are very low, with widened, short conical crowns. The dorsal surface of the dentary is widened, has a rugose thickening, which was probably covered by cornified tissue. Its surface forms a single pressing plane, which is compose of low crowns of the cheek teeth; in the caudal region, they are often arranged in parallel of two or three in a series. One maxillary tooth is increased and forms a deep fossa in the surface of the dentary. This is clear evidence of the absence of longitudinal grinding movements of the jaws. Wear facets on the lower incisors look like semicircular depressions, supporting the use of only vertical movements. The quadratum–articulare joint is considerably lowered; the postdentary part of the jaw is shortened, undoubtedly

increasing the pressure along the entire extent of the dental surface. In connection with this, the jaw condyles are convex on the quadrate and concave on the articular bone, the lateral margin of the quadratojugal (apophysis supracondylaris) and the medial surface of the surangular (fossa lateralis) have special areas, which prevent lateral movements of the jaw. It is possible that special structures restricting transverse movements are connected with a significant reduction of the crista transversalis of the pterygoid, which is shaped like a small tubercle that does not reach the lingual surface of the lower jaw. Streptostyly is impossible, since the fossa quadratica on the squamosal is flat, the capitulum quadrati has a flat, poorly expressed surface, and the lower part of the quadrate process is thickened, coming in tight contact with the QQJ-complex. It is noteworthy that, in this taxon, the lower margin of the vomer is only slightly elevated and the nasopharyngeal meatus is poorly developed. However, the medial margins of the maxillaries are widened and strongly overhang the choana. Thus, the two groups considered followed essentially different routes in solving the problem of isolation of the mouth and respiratory regions. Structural details of the sound-conducting apparatus are not known; however, the development of the fossa elliptica on the occipital surface of the squamosal, the presence of the sulcus stapedialis on the quadrate and a small thickened margin of the paroccipital process, which undoubtedly corresponds in topography to the postoccipital lamina, and the presence of laterally open postquadrate cavity suggest that these animals had a typical monocrural–postquadrate apparatus. The unique trend in the evolution of Anomodontia is represented by Suminia (Fig. 67c). The subapsid fenestra of this animal is developed to a much greater extent than in the previous groups, and its zygomatic arch is elevated above the level of the orbital center. The internal surfaces of the squamosal in the region above the quadrate and the entire lower surface of the zygomatic arch are turned obliquely laterally and open a large cavity for the external portion of the jaw adductor. Hence, the dentary has a special muscular area on the lateral surface of the dental ramus. However, the major character distinguishing Suminia from the taxa considered above is the formation of a convex condyle on the articular bone. The condyle is formed by a high crest (crista interarticularis) that enters a deep eminentia intercondylaris of the quadrate; the surfaces of condyles of the quadrate and respective fossae in the articular bone become flatter and form only the slopes of the new condyle. This design allows longitudinal movements of the jaw, as a high semicircular crista interarticularis as though rolls on the surface of the eminentia intercondylaris. This is a unique pattern of longitudinal jaw movements, because the high position of the posterior margin of the surangular, which is characteristic of all Eotherapsida and connected with the development of the ala angularis, makes mere anterior elongation of the PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

articular surface of the articular bone insufficient, although this often occurs in other, even very primitive groups (for example, Diadectidae, Bolosauridae). Suminia has an unusual dentition; however, it is easily derived from the dentition of, for example, Venyukovia. The incisors are only distinguished by less massive and narrower apices, they are shaped like elongated pointed blades, with well developed bordering lateroconules and narrow V-shaped borders on the lingual and labial sides. The cheek teeth also show minor distinctions, the only significant character is the very well developed labial and lingual cinguloids with a few large lateroconules. However, the last feature is only observed in very young crowns of replacement teeth; in Venyukovia or Ulemica, these stages have not been recorded. The major feature of the dentition of Suminia is the uniform plane of wear facets for all teeth in the row; this agrees with the hypothesis of the presence of longitudinal movements of the lower jaw. Since the first premaxillary incisor has a concave lingual wear facet and the first lower incisor has a convex labial facet, it is possible to conclude that, in the case of extreme protraction, the lower jaw did not extend beyond the first upper incisor. Suminia probably had a soft palate, this is supported by the preservation on the palate of plates covered with small teeth (see above). The grinding system of this type undoubtedly had many imperfections. In the presence of a very thin enamel layer, the teeth were worn rapidly and intensely up to the opening of the tooth canal; the jaws simultaneously contained up to three tooth generations. The teeth are undoubtedly disproportionately large relative to the skull size. In the case of this dental structure, an increase in body size was undoubtedly unfavorable in regard to energy consumption. Therefore, all Galeopidae are very small. In the other Anomodontia, the jaw teeth are reduced, at first, in the incisive region and, then, in the cheek region. Australobarbarus has several cheek teeth (Fig. 68a), which probably still functioned as a fence structure; in the others (Idelesaurus: Fig. 68b; Dicynodon: Fig. 68c; Delectosaurus, Elf, Vivaxosaurus), the teeth completely disappeared. The widened symphyseal regions of the premaxillae and dentaries have a porous surface, and the limbus dentalis has a more or less developed cutting marginal crest with the same surface. The function of teeth was probably performed by dense cornified jaws. This statement is supported by the increased nutrition of the jaw margins of the premaxillae, maxillaries, and dentaries; this is evident from the presence of many small foramina for blood vessels within these regions. Many taxa have well-developed maxillary tusks, usually with smooth, even wear facets on the lingual surface of the apex; this is probably accounted for by rubbing against a dense horny tissue. The cornified tissue covered the ventral surface of the premaxillae, participated in the formation of the lateral margin of the choana ossica and the dorsal surface of PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

963

the dentaries; however, morphologically differentiated horn plates, such as a horn beak, were not formed. Such plates would have formed distinct traces of porous growth grooves on the margins, as always occurs in the case of the development of pholidotic plates overlapping cranial bones (in lizards and turtles) or the formation of horn covers (osteodermal outgrowths on the skulls of Pareiasauria: Ivakhnenko, 1987; horn cover of Bovidae). The development and perfection of this apparatus for food treatment were undoubtedly connected with the basic features of the skull structure in this group. This primarily concerns a relative increase in volume of the cavity for temporal musculature. The region of the external portions, which is connected with the subapsid incisure, increased mostly by the elevation of the zygomatic arch and the external curvature of its internal surfaces. The region of the synapsid fenestra increased mostly by narrowing the parietal shield, up to the formation of a relatively narrow parietal crest (for example, Delectosaurus and many other taxa). The strengthening of muscles causes some essential characters of this group, such as tight fusion in the regions of the premaxillary and dentary symphyses, and constant presence of the interparietal, which reinforces the sagittal suture. The cornification of the palatal surfaces of jaw bones prevents the formation of the hard palate through the mode usual for many tetrapod groups, i.e., the development of special plates of the maxillaries and (or) palatines. The group considered achieved an unusual solution of this problem through a posterior expansion of the palatal plates of the premaxillae; this provided an increase in area of the grinding surface. In particular, in Australobarbarus and related primitive taxa, the anterior margin of the choana ossica is displaced close to the level of the midlength of the maxillaries; in Dicynodontidae, it is positioned at the level of the posterior edge of the maxillaries or even more posteriorly. At the same time, the palatal plates of the palatine and pterygoid are narrowed and curved dorsally, forming the lateral frame of the well developed nasopharyngeal canal. All East European representatives of Dicynodontidae are identical in general skull design and differ mostly in skull proportions, i.e., in the height-to-width ratio, elongation of the temporal or preorbital regions, and in the development of the parietal crest and the plates of the squamosal, which frame the temporal region. These differences are undoubtedly connected with certain functional features of the jaw system, and they should be analyzed using not only morphological but also engineering (biomechanical) approach of a sort. Among these taxa, Vivaxosaurus is of particular interest, since it shows a distinct ventral inclination of the preorbital region (the basipalatal angle is about 30°). The bend of the palatine plate is positioned close to the level of the pterygoid flanges, anterior to the region of the basipterygoid articulation; therefore, a klinorhinal braincase (like that of

964

IVAKHNENKO

(b)

(a)

(c)

Fig. 68. Skulls of Anomodontia: (a) Australobarbarus kotelnitshi Kurkin, 2000, holotype PIN, no. 4678/2; (b) Idelesaurus tataricus Kurkin, 2006, reconstruction based on holotype PIN, no. 156/114 and specimen PIN, no. 156/121; and (c) Dicynodon amalitzkii Sushkin, 1922, holotype PIN, no. 2005/38. Scale bar, 5 cm.

Dinocephalia) is not formed. An increase in klinorhiny is characteristic of Triassic groups, such as Lystrosauridae and Kannemeyeriidae, which are beyond the scope of the present study. Thus, the major processes in the cranial evolution of Anomodontia are connected with the reduction of jaw teeth, the formation and perfection of longitudinal movements of the lower jaw, and the development of the subapsid fenestra, transition to the difenestral skull design. The basal position in this lineage is probably occupied by taxa with complex tooth crowns; this follows from the structure of the juvenile teeth of Suminia. However, as indicated above, the use of these teeth, even in the case of primitive vertical movements, does not require the complex crowns of this type, and their elements are rapidly demolished by wear. Hence, the cheek teeth of the ancestral group functioned essen-

tially differently. Primitive groups, such as Venyukoviidae and Ulemicidae, used the cheek teeth in combination with vertical movements of the jaws and, then (at the level of organization of Galeopidae), developed a design with the horizontal protraction of the jaw. However, the dental structure was inconsistent with the intensification of food treatment; as a result, the teeth were reduced and replaced by cornified jaw covers. At the initial stages (about the level of Australobarbarus), the cheek teeth functioned as a fence structure, restricting laterally the mouth cavity; subsequently, they were completely reduced. The design produced was probably successful for certain conditions. Further slight changes were undoubtedly connected with searches for optimal solutions in particular adaptive trends; thus, they give no way of performing functional analysis at the modern state of understanding the group. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

CHAPTER 5. ECOBIOMORPHS Actually, the entire evolution of Eotherapsida was a process of natural development of their ecobiomorphs (Ivakhnenko, 2005a), which was mostly connected with changes in physiographic conditions of habitats. The biomorph (or life form after Bykov, 1983) determines the position of a particular taxon in community and relationships with certain biotopes. The ecobiomorph additionally determines the character of its obligatory ecotope. In extinct groups, the ecobiomorph is established through the reconstruction of the general morphological syndrome of a taxon. In addition, it is necessary to analyze every supplementary data, including the taphonomic association, proportions of taxa in the oryctocenosis, lithogenesis of localities, etc. Available data on the genesis of the majority of Dinomorpha localities of eastern Europe were provided in the previous study (Ivakhnenko, 2003c); in the present study, only necessary information on taphonomic associations are listed (see Chapter 1. Material). In my opinion, the ecobiomorph of any taxon is closely related to, and substantially determined by its morphological syndrome. The ecobiomorph is the only way to interpret certain features and must be included in the diagnoses of taxa. To date, the analysis of ecobiomorphs of Dinomorpha (as well as other primitive tetrapod groups) has not been developed satisfactorily in conceptual respect. At present, it is based mostly on approximately estimated relationships of known morphological features of the skeleton structure of extinct and extant groups and search for possible analogues. Certainly, this is the only accessible way, but this analysis is while restricted to mostly intuitional search for approximate general comparisons of body sizes and approximate proportions, similarity in the general pattern of dentition, etc. In my opinion, this analysis should be based on exact accounts of a number of determining parameters. They include the estimates of particular forces and their distribution in the skull, functional potentiality of the jaw system (not only structure, but also durability of teeth), strict accounts of locomotor possibilities, and even, where possible, estimation of functions of the sensory systems (olfactory, auditory, etc.). However, these data will probably be obtained in the future, possibly remote future. This is connected among other things with the fragmentariness of available fossil material, and an insufficient morphological understanding of the absolute majority of both extinct and extant groups. However, if we do not attempt to reconstruct modes of life at present, even though they are tentative and poorly grounded and do not use them in diagnoses, some interesting and important data may be lost. In particular, as different groups are referred to the same biomorph, we should search and estimate their structural similarities and distinctions, keeping in mind these relationships, and distinguish features connected with genetic relationships from similarities caused by a uniPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

965

form mode of life. In my opinion, the majority of structural features of Dinomorpha are caused by their biomorph evolution. All one has to do is to operate with caution with any reconstructed biomorphs and take them as working hypotheses that require further development and check. Our experience suggests that the lowest convenient taxonomic rank for this analysis is that of family group. Family-rank taxa are usually most resistant to various rearrangements and specifications of the taxonomic systems; in addition, detailed elaboration of biomorph models, intended for the establishment of taxa below family rank, sharply decreases the reliability of these reconstructions. Unfortunately, at present, it is possible to use only a very small proportion of characters, such as the body size, jaw structure, and dentition, for reconstruction of biomorphs. We are strictly limited in the data on the postcranial skeleton, these data are merely absent for the majority of taxa. To date, the general scheme of the evolution of biomorphs has been developed insufficiently. Regarding Dinomorpha, the general working scheme of relationships between morphology and biomorphs was proposed in the previous studies (Ivakhnenko, 2003c, 2005b, 2006); however, it requires further development and specification. Since the actualistic models provide vivid images, even if they are very approximate, it is worthwhile to consider extant analogues of extinct biomorphs whenever possible. These analogues are only very superficial and exist for a part of extinct biomorphs. They are well grounded mostly for predators, because their relationships with prey may be similar to what is observed in the modern fauna; however, they may not be extended to similarity in behavior. As for phytophagous groups, it is usually rather difficult to provide analogues with extant biomorphs because of essential differences between Paleozoic and Recent producers. The biomorph history of a group is usually closely connected with morphological evolution. In Dinomorpha, this history certainly began from the stage of the formation of all Theromorpha. This stage provided a number of important structural characters, which contributed considerably to the subsequent range of biomorphs accessible for Dinomorpha. In my opinion, the Theromorpha were formed as terrestrial (reptilian, obligatory terrestrial morphophysiological grade) derivatives of amphibian angustitabulars closely related to Gephyrostegida (Anthracosauromorpha) and retaining the apopareial design of the temporal region (Ivakhnenko, 2003c, pp. 352, 353). From this group, Theromorpha probably inherited features important for further evolution, such as a resonating periangular cavity of the lower jaw, a derivative of the groove of the infradental diverticulum of the seismosensory system of ancestors (Ivakhnenko, 2003c, pp. 354, 355). The position of the primitive sound-perceiving apparatus in the region of the lower jaw is probably connected with the sound conduction through the ground and, hence, suggests a very low initial position of the body (Tatarinov, 1958). This position of the body is not character-

966

IVAKHNENKO

istic of active xerophiles; consequently, the nearest common ancestor of Theromorpha was likely amphibiotic. Certainly, the adaptation of the initial auditory apparatus (compact skin in place of the operculum— hyomandibulare–stapes—foramen ovale—vestibular perilymph) with the use of resonating cavities is extremely primitive; however, primitive reptilian groups (such as Captorhinomorpha) lack special improving structures; therefore, at the initial evolutionary stage, even this adaptation could have provided Theromorpha with certain advantage. As the state of an obligatory terrestrial group was achieved, a unique integument was formed that probably initiated the development of the woolly cover of their mammal descendants (Ivakhnenko, 2005b, pp. 439, 440). The protection of the body against drying was a particularly urgent task at the moment of transition to obligatory life on land. The lissamphibian trend, involving the formation of mucous skin glands that dress the body in a natural reservoir of a sort, contrasts with the obligatory terrestrial mode of life. This cover retains moisture only slightly better than an open water surface; and, as the body grows, the evaporation surface sharply increases. The second trend involves the development of the scales covering the body. This is a pattern of great potential for all reptiles (Archosauromorpha, Lepidosauromorpha, Chelonia). At the same time, a continuous squamous cover stops transpiration and skin respiration and, hence, results in significant inner physiological changes and essential modification of skin. A comparison of the dermal structure and physiology of living reptiles and mammals (Tatarinov, 1976) conflicts with the hypothesis that the ancestors of mammals had a reptilian isolating scale cover. Consequently, theromorphs realized the third (mammalian) trend, the formation of an air-vapor layer by means of hair and sweat glands. Such a cover not only prevents excessive transpiration but also protects to a certain extent from overheating (Cowles, 1957). Certainly, I mean only the primary hair probably resembling the guard hair, which functioned as a porous cover regulating evaporation. Specialization of such guard hairs in certain areas of the body for tactile function resulted in the formation of the vibrissae (Brink, 1956) characteristic of primitive Theromorpha. The woolly–sweat cover allowed gradual improvement of physiological mechanisms of protection from drying. The improvement of the primary woolly cover, the formation of the undercoat (with many hairs, stretching from a common bulb) could have been connected with the development of the homoiothermy and, consequently, corresponded to the mammalian morphophysiological grade (Crompton et al., 1978). The entire history of the development of Eotherapsida was probably connected with the improvement of their physiology towards the mammalian pattern because, as they achieved a large body size, the air-vapor envelope could have resulted in overheating in a warm, humid environment. Active movements of a large animal with imperfect thermoregulation

probably require a large area for heat exchange. In particular, the dorsal vertebrae of primitive Sphenacomorpha (Sphenacodontia, Edaphosauria) had high neural spines, which formed a peculiar sail. The neural spines were densely twined by imprints of large branching vessels and were probably covered with a strongly vascularized membrane. In large Dinomorpha, such as Burnetiidae, Estemmenosuchidae, Anteosauridae, and all Tapinocephalida, the supraorbital and frontoparietal regions have extensive pachyostotic thickenings, which sometimes pass into high hornlike outgrowths. These formations are also twined by vessels coming from a special blood lacuna inside the frontal. The Gorgonopida, which were probably better adapted in this respect, have only small vascularized fields in the frontoparietal region, while, for example, the Dicynodontia lack a trace of special organs in this region (the same is true of Eutherapsida and, certainly, of mammals). These features of the syndrome, which were formed at the initial stages of the evolution of the group, had a significant effect on the mode of life of primitive Dinomorpha. Thus, it is possible to consider the major characters of the morphophysiological grade as a syndrome that determines biomorph potential of the entire Eotherapsida stem. Apparently, the physiology of the group was primarily advantageous for hygrophiles or, at most, mesophiles; therefore, they dominated the near-water ecotopes and prevailed among terrestrial groups in the localities investigated. As early as the Early Permian, the group included efficient obligatory phytophages (Edaphosauria); this was connected to a certain extent with the necessity to increase the body size. This is impossible at the amphibian morphophysiological level. At the same time, among the Theromorpha as a whole, up to the level of mammals, no groups connected with marine basins or air environments have been recorded. On the contrary, among reptiles with scale cover of approximately the same level, both marine and glinding taxa (Mesosauria, Weigeltisauria) appeared rather early; this is probably connected with the different physiology of primitive Theromorpha. Consequently, it is possible to regard Theromorpha as a group that reached the reptilian (obligatory terrestrial) morphophysiological grade without developing isolating scale cover (pilidosic group). The higher biomorphic divisions, i.e., the nanophagous, phytophagous, and predatory megabiomorphs, are usually reconstructed with confidence. The lower divisions, biomorphs and microbiomorphs, are determined by particular specialization trends and are connected with the morphophysiological grade of a group analyzed. In practical work, it is necessary to keep in mind that, within megabiomorphs, the boundaries between biomorphs are rather conditional in the overwhelming majority of cases; transitional variants are sometimes observed even between megabiomorphs. It is possible to designate the obligatory connection of a biomorph with a particular ecotope type (ecobiomorph) by respective ecotopes, as follows: hydrophiles (aquatic), PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

hygrophiles (amphibionts), mesophiles (near-shore–terrestrial, more or less connected with reservoirs), and xerophiles (inhabitants of dry areas located far from reservoirs). Among the Dinomorpha investigated, not only xerophiles but even pronounced mesoxerophiles are undoubtedly absent. This is probably connected with a certain physiological level of a group or results from an artifact because all presently known Late Paleozoic localities are connected with basins and, hence, provide respective taphocenoses. The nanophagous megabiomorph is probably most primitive, initial for the others. Formally, nanophages are predators; however, it is hardly expedient to assign them to this megabiomorph; thus, they are established as a separate megabiomorph. In the case of nanophages, prey is incomparable in size with the predator, and the latter is an active collector. Primitive nanophages of different groups consume various food objects (aquatic sclerophages, terrestrial entomophages, fossorial animals); however, all known Dinomorpha were probably obligatory or predominantly entomophagous. In this case, it is hardly probable that different taxa specialization in feeding on different objects; this occurs only in the case of feeding on colonial insects, which have not been recorded in the Paleozoic. In regard to ecotopes, nanophages can lead various modes of life (terrestrial, arboreal, aquatic ecobiomorphs); however, within Dinomorpha, only terrestrial representatives of the megabiomorph have been recorded. According to the definition of Dinomorpha, the basic feature distinguishing this taxon from Sphenacomorpha is the squamosal isolated from the quadratojugal and the formation of the streptostylic QQJ-complex (Ivakhnenko, 2003c, p. 358). The initial mobility of the complex probably provided certain advantage to small entomophages (lizard-like biomorph type). The Nikkasauridae, which display the most primitive morphology within Dinomorpha, have mobile lower jaw zones in the regions of the QQJ-complex, quadratum–articulare joint, and in the symphyseal region. This structure probably facilitated manipulation of small, active, and relatively firm prey (insects). The Nikkasauridae are small animals (with the skull no more than 5 cm long). They have a grasping, pointed anterior jaw teeth (of the centroconic type), the cheek teeth remain piercing. The development on the cheek teeth of supplementary projections (lateroconules) is probably caused by the ability of a more efficient treatment of relatively rigid and viscous chitin of insects, without jamming in it. It is noteworthy that these animals have sclerotic rings, with wide plates abruptly narrowing the aperture; may be this provided better vision. One of two genera examined (Reiszia) has a canine-shaped tooth in the upper jaw. This feature may be evidence of the development towards macrophagy; however, lizards (some agamas and iguanas) also have increased upper teeth, although they do not turn to obligatory macrophagy. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

967

Judging from the dental structure, the Microuraniidae were much more specialized entomophages. They have canine-shaped piercing teeth in the upper and lower jaws, which are equipped with thickened cinguloid at the crown base. The cheek teeth are massive relative to their size, with a pointed apex and distinct cinguloid. The incisors are unusual, with a widened and flattened occlusal surface but without wear signs. The character shared by all Dinomorpha is a tendency of predominant use of the incisive region of the jaw as the working area; this design of incisors of Microuraniidae could have been used for shattering the strong rigid chitin (for example, of beetles). The fact that all Dinomorpha, even the most specialized groups, passed the stage of entomophagy is supported by the initial complex crown structure of the cheek teeth, which seems unnecessary for functioning. A vivid example is provided by the teeth of primitive Anomodontia. The complex structures of the tooth crown described above in Suminia were rapidly worn out in functioning teeth even in juveniles; this suggests that they were not acquired for grinding firm food. They were probably formed in an ancestral group specialized in entomophagy and splitting the chitin of insects. It is hardly probable that small entomophages were an initial group for other megabiomorphs. This is primarily connected with the necessity of the preservation of relatively small size, corresponding to feeding on small insects. However, when feeding on insects and their larvae, which occurred in vegetative substrate (for example, fructifications, fruit, fungi, etc.), animals could have turned to facultative phytophagy (Hotton et al., 1977). The initial stage of this process, i.e., feeding on any food objects of suitable size, ranging from small invertebrates to the most nutritious parts of plants, could have resulted in omnivory of a sort. However, the majority of nanophages, except for very narrowly specialized forms, are more or less omnivorous, consuming everything edible. Therefore, it is hardly expedient to establish this stage as a separate biomorph. At the same time, transition to feeding on plants requires certain increase in body size in connection with the necessity of lengthening the alimentary tract. This causes an increase in size of predators because, in primitive groups, predators that are somewhat larger than prey are most efficient. An increase in size results in transition from small insects to obligatory predation or obligatory phytophagy. This probably gave rise to the initial stages of phytophagous and predatory megabiomorphs. In Dinomorpha, these processes were marked by different directions of specialization of the dental apparatus. The predatory megabiomorph of Dinomorpha is distinguished from other primitive tetrapods by a unique feature—alternative cutting system of the upper and lower canines. The flattened falciform canines along with a semicircle of incisors form an apparatus cutting pieces out of the body of prey that are suitable for swal-

968

IVAKHNENKO

lowing. This dentition is not suitable for cutting rigid skin covered with scales; hence, it was probably formed as a result of feeding on the pilidosic Theromorpha. The development of this jaw design was probably an important apomorphy of predatory Dinomorpha. Other primitive predators of various taxa usually have only increased teeth in the maxillary, without specialized antagonists; they were only intended for participation in killing prey, which is smaller than these predators. More highly specialized groups (for example, Sphenacodontidae or primitive Proterosuchidae) combine large canine-shaped teeth in the maxillary with strengthened anterior incisors in the upper and lower jaws, which form pincers participating in tearing large prey with a relatively soft skin. Prey was rather homogeneous in physical properties and differentiated mostly in behavioral features, which are difficult to reconstruct in our analysis. Thus, some general features of representatives of the megabiomorph in the groups analyzed are recognized. Predatory Dinomorpha are in general characterized by relatively weak jaw muscles. This is probably connected with the specific structure of the dentition (see above). The development of alternative canines and flattened incisors with sharp cutting borders equipped with a row of small sharp lateroconules, which form an apparatus for cutting out pieces of prey body, undoubtedly provided a considerably weaker compression of jaws than in the case of a tearing apparatus. The majority of these predators were probably not very specialized with reference to prey, which comprised almost any animals of suitable size, whereas the size-class of prey was probably most important. These predators could hunt only prey comparable with them in size. Their dentition prevented hunting small prey (in contrast to modern predators, e.g., wolf or fox, which occasionally hunt mice). In this case, the set of predators corresponding to the diversity of phytophages was likely based only on the diversity in sizes. It is possible to hypothesize that there were many predatory biomorphs slightly differing in morphology and distinguished mostly by size. The fact that presently known taxa mostly belong to large and medium-sized classes is attributable to various factors. First, as mentioned above, relatively small solitary predators rarely occur in burials. However, this may be connected with unknown structural (or physiological) features of primitive animals, which prevented the development of an efficient small-sized predator. If this is the case, the role of this biomorph, which is necessary in any full-scale community, could have been played by young individuals of large specialized taxa. This is indirectly supported by the fact that, in some groups of primitive predators, different ontogenetic stages show differences in dental structure, in particular, in canines (Bakker, 1982; Ivakhnenko, 1999). In all known predatory Dinomorpha, the visual analyzer distinctly prevailed. The orbits are usually disproportionately large compared to the skull size. The frequently well-developed sclerotic ring could have been

connected with good accommodation of eye. The mobile elements of the ring could somewhat compress the anterior part of the eye ball, contributing to focusing. At the same time, it is hardly probable that the olfactory system was well developed. In the majority of known taxa, the posterior part of the relatively small nasal cavity is occupied by the mesethmoid structures, the anterior part contains a well developed cavity of the intermedial process of the septomaxillary, which contained the lateral nasal gland, and the middle part is occupied by a large canine capsule and the maxillary sinus (antrum of Highmore) located posterior to it. In addition, taking into account probable penetration of the nasopharyngeal lining, ascending high on the lateral walls of the internasal septum (which was recognized in Biarmosuchus, see above), too little area remains for the development of the olfactory epithelium. Possibly, the auditory apparatus, which was rather unusual in Dinomorpha, was relatively weakly used for hunting. Perhaps, the repeatedly indicated differences in volume and configuration of the periangular fissure even in closely related taxa suggest resonator amplification of a relatively narrow acoustical spectrum rather than welldeveloped hearing with a wide range of auditory signals. This probably means mostly selective recognition of sounds produced by individuals of the same species and connected with reproduction or protection of habitats. The biomorph of small terrestrial predators–generalists (marten-like biomorph type) is very poorly understood. This is not surprising in regard to medium-sized predators, which rarely occur in the burial. Some representatives of Ictidorhinidae are probably most similar in morphology to these forms. Unfortunately, this group is poorly represented in the collections examined (Ustia, Biarmosuchoides) and poorly understood. These medium-sized forms are undoubtedly not very specialized predators, most likely with a very wide range of feeding objects. The alternative canines are well formed, relatively small in size, somewhat thickened. The cheek teeth are weakly specialized, of a primitive long-leaf type. It is possible that the known Ictidorhinidae were mostly euryphagous rather than true predators. At the same time, collections include many bone fragments and isolated teeth, which probably belong to small predatory Dinomorpha; however, the taxonomic position of these specimens is uncertain (they are excluded from the material of the present study). Perhaps, this biomorph was in fact more diverse in taxonomy than is presently thought. The larger biomorph (doglike biomorph type) is much better understood. It comprises representatives of various taxa, such as Alrausuchidae, Phthinosuchidae, and primitive Gorgonopidae. These taxa have typical cutting alternative canines and cutting incisors. In a number of taxa (Alrausuchus, Dinosaurus, Viatkogorgon), the cheek teeth are thickened, slightly curved, with well-developed cutting borders equipped with many small sharp lateroconules (carnassial type); they probably participated in food treatment. A particular PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

biomorph (tigerlike biomorph type) is composed of very large-sized forms (such as Kamagorgon, Leogorgon, and Rubidgeidae that resemble them in morphology), with a short and massive skull. The infrequent occurrence of these groups in oryctocenoses corresponds to the high level of terrestrial adaptation, probably, mesophilic life style. A special ecological group is formed of the higher Gorgonopidae (Pravoslavlevia, Sauroctonus, Suchogorgon, etc.). They show a well-pronounced streptostyly of the skull, with much better developed jaw muscles. The canines are very long relative to the skull size, flattened and thin, slightly curved, with well-developed serrated borders; they are only suitable for killing prey and, along with cutting incisors, for cutting out large pieces of food objects. The postcanines are very small, only slightly occluding; they undoubtedly did not participate in food treatment. The inclined position of postcanines in the jaws, along with the presence of kinetic zones between the dentary and postdentary part of the lower jaw, probably facilitated the swallowing of large pieces. This suggests that prey was much larger than, or at least comparable in body size, to predators. The skeleton structure suggests that they were active terrestrial animals (Colbert, 1948; Hotton, 1980; Kemp, 1982; etc.); at the same time, they are more frequent in oryctocenoses than the previous groups. The taxon includes many related forms with minor morphological distinctions of uncertain taxonomic significance (Ivakhnenko, 2005b); in some cases, they are probably of population level. This suggests that we for the first time in the Late Paleozoic deal with a multimorph predator group, which formed small populations and hunted relatively large prey. It is hardly probable that at such a primitive developmental level it is expedient to speak about a pack of animals in the sense that generally applied to living dogs or wolves; however, certain elements of such social behavior could have been achieved. Perhaps, this was related to unique among Dinomorpha changes in the structure of the sound-conducting apparatus of Gorgonopidae under study, which lost the resonator cavity and acquired a very unusual large “tympanum” formed of almost the entire postdentary part of the jaw (Ivakhnenko, 2005b, p. 442). This design could have appeared because of the necessity to perceive a wide range of sounds, which was connected with considerable changes in behavior. Representatives of Burnetiidae, such as Proburnetia and Niuksenitia, are similar in appearance to doglike predators. These animals show approximately the same body size; however, the presence of pachyostotic hornlike outgrowths on the skull arouse some suspicion, since they are atypical for predators. The canines are relatively short, massive, with weak cutting borders; the cheek teeth are short and somewhat thickened. A high nasopharyngeal meatus restricted by distinct narrow toothed palatal tubercles is well developed, although it is not characteristic of predators. Based on the assumption concerning the origin of this group from primitive PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

969

predatory Gorgonopia, it is possible to conclude that they underwent a respecialization of a primary predator into an omnivore, showing transition to a piglike biomorph type. A special biomorph, which has no analogue among extant taxa, is represented by members of the families Biarmosuchidae, Anteosauridae, and Inostranceviidae. They are very large animals, with at least 3-m-long body. They have a well-developed cutting alternative dentition, with very long flattened (saber-shaped) canines, probably adapted for a very thick soft skin of prey. The most important structural features of these animals are a relatively large, narrow, and high skull with considerably elevated nares and orbits, and thin, gracile skeleton with wide manus and poorly ossifying epiphysial regions of tubular bones. Representatives of these groups are abundant in some localities and differ in individual age. As in the groups described above, the dental structure suggests hunting prey of approximately the same size as predators, and the postcranial skeleton corresponds to the aquatic mode of life. These animals could have been hygrophilous macrophagous predators tightly connected with reference to trophic specialization to very large aquatic or amphibiotic phytophages. As mentioned above, young individuals could have played in community the role of medium-sized amphibiotic predators–generalists. This agrees with a significant disproportionate elongation of the canine, for example, in the ontogeny of Biarmosuchus (Ivakhnenko, 1999). Apparently, the study of the postcranial skeleton of these interesting animals would improve considerably the knowledge of the mode of life of various representatives of this biomorph. In particular, even a superficial comparison of the forelimbs of Inostrancevia and Biarmosuchus shows essential structural differences. The first has wide, flattened ungual phalanges, while the second displays high and narrow phalanges, with the distal ends curved abruptly downwards and well-developed flexor tubercles. In any event, these differences may be connected with certain distinctions in the mode of life and hunting; however, a better understanding of their role requires additional examination. A biomorph even more closely connected with water basins is probably represented by the Archaeosyodontidae and Syodontidae (Microsyodon, Archaeosyodon, Syodon). In the evolution of these groups, the formation of a special canine of the fishhook type from the increased cutting tooth characteristic of primitive predators is clearly traced (Ivakhnenko, 2003c, pp. 377, 378). This “canine” is a slightly conical cylinder positioned inclined anteriorly in the jaw, with the apex curved sharply posteriorly. This “canine” is not adapted for killing or dissecting large prey, since in the case of a vertical stress (which is typical for cutting or tearing canines), this stress falls on the curvature rather than on the pointed apex. This canine structure is adapted for capturing prey that is small relative to the size of predator and covered with scales, i.e., fish or labyrinthodonts. This agrees with the evolution of cheek teeth

970

IVAKHNENKO

from strengthened holding (Archaeosyodontidae) to rounded pressing (Syodontidae), which are adapted for crushing relatively rigid prey. In this case, we probably deal with an otter-like biomorph type of a sort. The hydrobiont mode of this group is indirectly supported by almost complete absence of thermoregulatory structures (which lose significance in aquatic environment) in the frontoparietal region of the skull roof of these very primitive taxa. It is interesting that Syodon has a very well-developed structure connected with an increased cerebellum, i.e., the pocket fovea floccularis of the greatest size relative to the braincase among all known Dinocephalia. This structural feature is probably connected with active movements in the threedimensional space, i.e., with an aquatic mode of life. The phytophagous megabiomorph is represented in Dinomorpha by many diverse biomorphs. Vegetation provides extremely rich food resources on land, and progressive development of phytophagy was a general trend in the evolution of this group. In contrast to predation, feeding on plants is connected with the treatment of matters with widely ranging physical properties; therefore, phytophages are much more diverse in morphology than predators. The only common trends are probably an increase in body size and strengthening of jaw muscles. The formation of obligatory phytophages followed from an increase in the proportion of plant component in the diet of primitive omnivorous taxa. In the Late Paleozoic, with specific plant composition, the primary phytophagy could have been connected with various relatively nutritious and easily digested components, such as young shoots, fructifications, other generative organs, and roots. This level of phytophagy was named frugivory (Ivakhnenko, 2005b, p. 457). This term comprises all groups of primitive phytophages before reaching the level of feeding on coarse fibrous plants. At the same time, various frugivorous groups, including not only Dinomorpha (Diadectidae, Niaftasuchidae, Venyukoviidae, Ulemicidae, Caseidae, Bolosauridae, Belebeyidae), also have some features in common (Ivakhnenko, 2005b). As the biomorph was recognized, it was indicated that its representatives are usually medium-sized, have strengthened chisel-shaped incisors and a relatively high nasopharyngeal meatus or even a more or less developed hard palate. The cheek teeth are relatively diverse in structure, in accordance to the extent of primitive food treatment characteristic of each group. The development of elements of the hard palate, grinding or chopping teeth suggests gradual transition to feeding on coarser fibrous vegetative parts of plants. The most primitive frugivores known within Dinomorpha are represented by the Niaftasuchidae. They are medium-sized animals, with a relatively weakly developed nasopharyngeal meatus and the incisors inclined strongly anteriorly with strengthened crests on the lingual surface. The self-sharpening of apices by incisors antagonists is very poorly pronounced; apparently, incisors were used mostly for tearing off food

objects (maybe fructifications). The cheek teeth are very primitive, with somewhat widened crowns of the short-leaf type; they probably played only a role of a border of the mouth cavity (fence structure). The Venyukoviidae (Otsheria, Venyukovia) probably represent a somewhat more specialized microbiomorph. The self-sharpening incisors have sharp anterior margins and are intended for cutting; the cheek teeth have distinct vertical wear facets produced by antagonistic teeth; this intensified their cutting ability. Since the jaw movements of these animals were restricted to the vertical plane, the dentition was probably used for a peculiar shearing of plant parts. The dentition of the captorinomorph Belebey functioned similarly (Ivakhnenko and Tverdokhlebova, 1987); hence, this was not a unique adaptation, and the study of Late Paleozoic vegetation will provide a better understanding of this topic. The Ulemicidae probably represent a similar but distinct microbiomorph, which combined vertical jaw movements with almost the same sharpened incisors, but short and rounded cheek teeth. One upper cheek tooth is considerably increased, but has the same rounded conical shape. In addition, the bone surfaces bearing teeth are widened and have pachyostotic thickenings, such that the dentaries have pits for the largest upper teeth, particularly for the increased tooth. This dentition is only suitable for crushing. Sclerophagy (i.e., feeding on, for example, mollusks or arthropods with firm armor) in this case is highly improbable, since it contrasts with the development of the enlarged upper tooth and widely spaced other cheek teeth unsuitable for efficient crushing individual and relatively small objects. This suggests grinding by repeated compression (chewing of a sort) of relatively large and firm plant fragments cut off by incisors. It is possible to regard these changes in the jaw apparatus as a primitive adaptation to initial herbivority. The Rhopalodontidae (Rhopalodon, Phthinosaurus) could have been primitive frugivores. Unfortunately, this family is only represented by jaw fragments. The structure of incisors is not known; however, judging from the size of alveoli, they were very small. The upper and lower jaws have distinct, although relatively small canines. However, the canines are not shaped like a thin cutting sickle (which is typical for predators). They are massive, circular in cross section, reinforced by longitudinal edges and lack cutting borders. Judging from the structure of alveoli, they were slightly inclined anteriorly and could have played the same role as increased incisors of the frugivores described above. The cheek teeth retain primitive (short-leaf) structure and lack a trace of wear; thus, they probably functioned only as a fence structure. This group probably evolved from a primary predatory biomorph. The Galeopidae are similar to frugivores in the general structure of the jaw system, with increased, inclined, and self-sharpening incisors. However, their lower jaws were capable of longitudinal movements, PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

providing sufficient amplitude for grinding effect. This is evident from the presence of a single longitudinal plane of wear on the cheek teeth. The cheek teeth are heavily worn, sometimes to the collum dentis and tooth canal. Worn teeth were rapidly substituted by replacement teeth, such that a jaw sometimes has up to three tooth generations. The palatine plates of the premaxillaries are posteriorly elongated, displacing posteriorly the posterior margin of the bony choanae, which are positioned high above the palatal plane. This mechanism was probably intended for the treatment of relatively coarse fibrous plants (Rybczynski and Reisz, 2001). The strong limbs of these animals, with broad feet and long digits, suggest that they could have inhabited swampy shores (Ivakhnenko, 2003c, p. 424). However, if this is the case, it remains uncertain which type of plant diet caused such a heavy tooth wear. Perhaps, an insight into the mode of life of these unusual animals may be provided by the following facts. The Kotelnich locality (collection PIN, no. 2212, Suminia getmanovi) has yielded several tens of skeletons and many isolated specimens, mostly durable teeth. The good preservation of complete skeletons suggests subautochthonous or even autochthonous burial with reference to this group. The trunk is relatively short, and the body size is very small for a typical frugivore, the skeleton is at most 30 cm long. These features combined with strong, long manual digits equipped with well-developed ungual phalanges suggest that the group possibly represented a digging (subfossorial) biomorph. The strong tooth wear could have been caused by feeding on plant roots covered with abrasive ground particles. It is hardly possible that these animals were aquatic or amphibiotic, and the large number of their skeletons in the locality considered is probably accounted for by a colonial mode of life. A peculiar biomorph that has no analogue among living animals is represented by the largest phytophages (Ulemosaurus, Estemmenosuchus). These groups were determined as original saprophagans feeding on partially decayed plants (Ivakhnenko, 2003c, pp. 390, 414; 2005b, p. 458). The saprophagous stage probably played a significant role in the gradual formation of the true phytophagy–herbivority (Hotton et al., 1977). The partially decomposed dead plants could have been the initial source of microflora necessary for assimilation of coarse fibrous plants. The existence of large shallow ponds covered with mats of dead tracheophytes, which turned into fermented silage of a sort, was quite possible in conditions of the Late Paleozoic (Ivakhnenko, 2003c). This feeding mode provided a considerable increase in body size in connection with the elongation of the alimentary tract. At the same time, it did not require significant specialization in the structure of the jaw apparatus and was accessible for primitive representatives of different groups. Among the taxa investigated, two, Estemmenosuchus and Ulemosaurus, were giants. These genera differ in the structure of dentition but are undoubtedly similar in its function. The relaPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

971

tively soft food mass was grasped mostly by the incisive region, which, however, was not specialized for cutting. The small, numerous cheek teeth (of the pronounced short-leaf type), with transversely broadened crowns, virtually did not occlude, playing only a role of a fence structure. The nasopharyngeal meatus and mouth cavity are relatively poorly isolated; this suggests rapid swallowing of plant matter, almost without treatment in the mouth cavity. Both taxa show extremely developed klinorhiny, which retained a high position of orbits combined with a lowered incisive region, which is necessary for feeding on food objects located low relative to the head. The “thermoregulatory” structures along with pachyostosis are very well developed. A shared feature is abrupt reorganization of the sound-conducting apparatus, transition to the postquadrate system and a decrease in the role of the resonator cavity, with simplification and narrowing of the periangular fissure and an increase in thickness of the ala angularis. These changes in large phytophages could have been connected with a need for perception of a wider sound range than that provided by the resonator system. As the biomorph was formed, the dentition of Estemmenosuchidae required significant morphological changes, since the group evolved from a primarily predatory ancestor. In particular, the cutting alternative canines transformed into tusks, circular in cross section and lacking cutting borders. The initial stages of these changes fell on the level of frugivores, i.e., Rhopalodontidae. The tusks and strengthened incisors deviate considerably labially and compose a grasping apparatus (Fig. 27b). No structures connected with primitive food treatment are observed. The dentition of Ulemosauridae is changed relative to primitive frugivores (Niaftasuchidae) to a much lesser extent, although it is much more specialized than in Estemmenosuchidae and provides certain food treatment. The massive broadened incisors of Ulemosauridae have distinct wear facets. The wear facets are shaped like flat depressions on the upper teeth and are flattened convexities on the lower teeth; this suggests the presence of primitive compression without true grinding of food objects in this region. Apparently, the strong development of the postorbital portions of the jaw muscles was also connected with an increased pressure on the anterior teeth, which was necessary for this. It is interesting that the upper jaw of Ulemosaurus has acquired increased canine-shaped teeth, which probably contributed to the isolation of an optimum volume of food for swallowing, as the tusks of Estemmenosuchus. It is possible that difficulties with the consumption of unusual food, such as dead plants, resulted in the formation of a special biomorph represented by Deuterosauridae. These animals are as large as saprophagans, but their dentition is changed considerably. The high, massive, and short skull had even stronger jaw muscles than in Ulemosaurus; the incisors are sharper, with less developed cinguloids; the increased upper tooth corre-

972

IVAKHNENKO

sponds in function to canine and is opposed to an increased canine-shaped lower tooth. The cheek teeth have become shorter and more massive, with rounded pressing crowns. This jaw apparatus probably developed because of transition to the secondary omnivory, scavenging, and nonspecialized carnivory. It is probably possible to designate this biomorph as a bearlike type. In almost all ecotopes, most of the biomass is usually composed of vegetative parts of plants; therefore, the major biomorph of terrestrial tetrapods is the herbivorous biomorph. In modern communities, they usually compose the basis of the tetrapod association of consumers of the first order. However, the assimilation of coarse fibrous vegetative food and establishment of the biomorph of efficient herbivore required essential changes in a number of structures related to the alimentary system, because Dinomorpha initially belonged to the biomorph of nanophagous predators. Since this work is devoted to the analysis of the skull design, we should first take into account the systems responsible for crushing plants in the mouth cavity. This is a special dental system (or a system efficiently replacing it), the design of the articular region, which provides longitudinal jaw movements of a sufficient amplitude, and improvement of the air-conducting system for the possibility of long food treatment in the mouth cavity. However, these changes in Dinomorpha were connected with certain difficulties. This particularly concerns the cheek teeth, which in all Dinomorpha have a minor contribution to the food treatment because of their structure. It is interesting that isolated incisors and canines are more frequent in localities than cheek teeth. The only exception is the cheek teeth of Suminia, which are probably more durable. At the same time, the dentition of Galeopidae, including repeated replacement of relatively large teeth, undoubtedly corresponds to small-sized frugivores, which consumed little amount of food rich in calories, and is insufficient for a large herbivore. Therefore, in herbivorous Dinomorpha (of which I examined representatives of Pristerodontidae and Dicynodontidae), cheek teeth were replaced by the cornified jaw cover, which formed an efficient, constantly growing rhamphotheca. The quadratum–articulare joint of two convex condyles, which was acquired at the morphological level of Galeopidae, is improved; the palatine plates of the premaxillaries expand posteriorly, displacing the pars posterius of the choana ossica (naria interna) from the region of food treatment. The strong portions of the jaw adductors were similar in function to the masseter of mammals and provided a strengthened pressure of jaws and a wide amplitude of their reciprocal movements (Crompton and Hotton, 1967). Perhaps, an important point is that, in contrast to the primitive groups described above (Dinocephalia, Gorgonopia), the higher Anomodontia lacked special thermoregulatory structures, and these groups were probably characterized by much more advanced physiology. This design possibly began to develop at the level of the subfossorial frugivore, the teeth of which were

replaced by cornified jaws, and strengthened canineshaped upper teeth (rather than incisors) were used for digging ground. Primitive Eodicynodon, which already developed all necessary structures, is similar in skull morphology to this hypothetical group; however, the small size is hardly compatible with the assignment of this animal to efficient herbivores. Certainly, the increased maxillary tusks were acquired for digging ground; this is supported by the constant growth of these teeth and self-sharpening by rubbing against the horn cover of the lower jaw. They probably had the same function in many groups (Cox, 1959); however, some other groups could have retained the tusks only as a sexual character (Sullivan et al., 2003). Among many isolated tusks, the collection investigated includes definitive (cylindrical) specimens, which have uneven wear traces produced by ground on the labial side of the apices. However, some tusks lack wear traces on the labial surface and have only a smooth lingual “mandibular” area (as is shown in Fig. 9b). The solution of this question requires a more thorough study of definitive tusks of various taxa. This design of herbivores was rather efficient in conditions of Late Paleozoic ecotopes; it gave rise to many microbiomorphs. The taxa examined (Australobarbarus, Idelesaurus, Dicynodon, Delectosaurus, Elf, Vivaxosaurus) are identical in general morphological pattern; the differences in skull structure are insignificant and, at the present state of knowledge, seemingly lack correlation with differences in the mode of life. Perhaps, in this very multimorph group, they were determined by certain features of various objects of obligatory feeding. It is possible that some extremely deviating groups, which are unknown in eastern Europe, belonged to some other biomorphs. For example, Myosaurus is believed to be insectivorous (Cluver, 1974) and Kawingasaurus is probably subfossorial (Cox, 1972). CHAPTER 6. DISTRIBUTION AND GEOGRAPHICAL RANGES The distribution in time and space is an essential parameter of any taxon. Regarding extinct taxa, this knowledge is always incomplete; this particularly concerns early tetrapods, which are usually represented by individual finds. These records usually only mark the presence of particular taxa in a certain time and certain point, but almost always does not correspond to the true time of appearance or disappearance or true geographical range. This particularly concerns the distribution of taxa in time. In my opinion, the time of existence played a very important role, since both appearance and extinction of a taxon was determined by the period of existence of particular ecological conditions. Certainly, the geographical range is also connected with the area of suitable conditions. This provides at least approximate reconstruction of temporal and spatial ranges of each taxon, using the material of associated taxa of PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

other tetrapod groups, that is, data on faunal assemblages. Faunal assemblages correspond to certain developmental stages of the territory, which are connected with changes in physiographic conditions (in regard to eastern Europe, see Ivakhnenko, 2001, 2003c). The greatest changes were undoubtedly caused by global factors; this allows the correlation of faunal assemblages of remote regions and revelation of possible migration routes of the tetrapod groups under study (Ivakhnenko, 2005a). Faunal assemblages are reconstructed based on the analysis of oryctocenotic associations; therefore, the time of their existence more or less closely correlates with the units of the biostratigraphic scale and, hence, the scale of absolute time. However, the biostratigraphic chart of the region studied and correlation with the general chart is based not only on the data on tetrapods. Therefore, in practical work, there may be cases in which comparisons of tetrapod assemblage with the data on other territories results in conclusions that distinctly contradict its biostratigraphic position. For example, this is possible for assemblages from refuges, which are connected with long preservation of certain habitats within local areas. Biostratigraphic questions are beyond the scope of the present study; therefore, the time and period of existence of particular taxa are determined here based on faunal assemblages including them (for biostratigraphic aspect of the problem, see Golubev, 2005). It is evident that the geographical range of a taxon is also connected with the distribution of suitable conditions. Although the tetrapod assemblages from the Late Permian of eastern Europe and South Africa are externally similar in taxonomic composition, a thorough comparison shows certain differences (Ivakhnenko, 2005a), which are undoubtedly connected with differences in physiographic conditions between these territories, which at the beginning of development were not only on opposite slopes of the Tethys Ocean but also on different sides from the equator (Olson, 1979). It is possible that further study of tetrapods from the Late Permian of Central Asia (China) will show one more isolated spatial range; perhaps, the same is true of southern Asia (India), the tetrapod assemblages of which are presently poorly understood. Smaller ranges are restored with certainty only in exclusive cases. For example, in eastern Europe, this concerns the Mezen and Ocher assemblages of approximately the same geological age, which undoubtedly occupy lateral positions in space (Ivakhnenko, 2001, p. 166). The analysis of the Mezen Assemblage shows that it is very primitive, probably representing a refuge with a long preservation of conditions of a broad marshy area along a flattened western slope of the Kazanian Sea. The Ocher Faunal Assemblage distributed on the eastern slope, along the Paleo-Ural Mountains differs considerably in taxonomic composition. According to different ecotopic conditions of these areas, the ranges of taxa of Dinomorpha are also different. The Isheevo Faunal Assemblage covers the southern and southeastern parts of the PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

973

East Europe Placket, while the Sokolki Assemblage occurs in the entire area of the placket (see Ivakhnenko, 2003c, text-fig. 54). The Kotelnich and Vyazniki assemblages have only been recorded in isolated localities, and their ranges remain unknown. The Mezen Faunal Assemblage is most primitive in the composition of Dinomorpha in eastern Europe. It comprises the following taxa: Parareptilia (Procolophonida), Nyctiphruretidae Efremov, 1938: Nyctiphruretus acudens Efremov, 1938. Parareptilia (Pareiasaurida), Nycteroleteridae Romer, 1956: Nycteroleter ineptus Efremov, 1938 and Bashkyroleter mesensis Ivachnenko, 1997. Parareptilia (Pareiasaurida), Tokosauridae Tverdochlebova et Ivachnenko, 1984: Macroleter poezicus Tverdochlebova et Ivachnenko, 1984. Parareptilia (Pareiasaurida), Lanthaniscidae Ivachnenko, 2007: Lanthaniscus efremovi Ivachnenko, 1980. Ophiacomorpha (Ophiacodontia), Varanopidae Romer et Price, 1940: Mesenosaurus romeri Efremov, 1938 and Pyozia mesensis Anderson et Reisz, 2004. Ophiacomorpha (Casesauria), Caseidae Williston, 1912: Ennatosaurus tecton Efremov, 1956. Theromorpha (Dinocephalia), Alrausuchidae fam. nov.: Alrausuchus tagax (Ivachnenko, 1990). Theromorpha (Dinocephalia), Niaftasuchidae Ivachnenko, 1990: Niaftasuchus zekkeli Ivachnenko, 1990. Theromorpha (Nikkasauria), Nikkasauridae Ivachnenko, 2000: Nikkasaurus tatarinovi Ivachnenko, 2000; Reiszia gubini Ivachnenko, 2000; and R. tippula Ivachnenko, 2000. Lepidosauromorpha (Eolacertia), Lepidosauria fam. indet.: Lanthanolania ivakhnenkoi Modesto et Reisz, 2002. This assemblage has no direct analogues in other territories. However, it possibly corresponds mostly to the primary tetrapod community of coasts of large basins (pioneer community: Ivakhnenko, 2006). If this is the case, the biomorph composition of communities of various areas was approximately the same, including small primitive nanophages, predatory and phytophagous omnivores of both amphibian and reptilian morphophysiological grades. The highest levels of specialization were probably represented by frugivores and predators–generalists. The taphocenosis of localities of these communities sometimes includes hydrophiles, i.e., various groups of ichthyophagous labyrinthodonts. The absence from the Mezen Assemblage of fish and tetrapods of the primary ichthyophagous food chain most likely reflects the specificity of the basin (Ivakhnenko, 2001), which was connected with the formation of available localities. The taxonomic composition of assemblages corresponding to these communities is determined by geographical and temporal ranges. The Mezen Assemblage probably reflects the developmental stage of communities corresponding to the initial penetration of primitive Dinomorpha. In addition to

974

IVAKHNENKO

Nikkasauridae (very primitive nanophages), Dinomorpha are represented in the assemblage by two groups of Dinocephalia. It is possible to regard Niaftasuchidae as a group similar in morphology to the basal Dinocephalia, whereas the Alrausuchidae are probably early native derivatives, which appeared because of deficiency in medium-sized predators–generalists in this assemblage. This suggests to regard it as an endemic assemblage; however, the presence of Caseidae, which are also widespread in the Early Permian of western Europe and North America is evidence that analogous assemblages are distributed wider, although they have not yet been investigated. On the other hand, the analysis of primitive groups in later assemblages of South Africa shows (Ivakhnenko, 2005a) that its primary communities included primitive Gorgonopia rather than Dinocephalia, as in the Mezen Assemblage. At the initial developmental stage, the faunal assemblage of the eastern coast of the Kazanian basin, i.e., the Ocher Assemblage, resembles the Mezen Assemblage. The initial subassemblage, the Golyusherma Faunal Subassemblage includes the following taxa: Batrachomorpha (Edopiformes), Melosauridae Fritsch, 1885: Melosaurus uralensis Meyer, 1857; M. kamaensis Gubin, 1991; M. compilatus Golubev, 1995; M. platyrhinus Golubev, 1995; and Koinia silantjevi Gubin, 1993. Batrachomorpha (Edopiformes), Platyoposauridae Huene, 1931: Platyoposaurus watsoni (Efremov, 1932). Batrachomorpha (Dissorophoidea), Dissorophidae Boulenger, 1902: Alegeinosaurus sp. Parareptilia (Discosauriscomorpha), Leptorophidae Ivachnenko, 1987: Leptoropha talonophora (Tchudinov, 1955) and Biarmica tchudinovi Ivachnenko, 1987. Anthracosauromorpha (Gephirostegia), Enosuchidae Konzhukova, 1955: Nyctiboetus kassini (Tchudinov, 1955). Captorhinomorpha (Captorhinida), Captorhinidae Case, 1911: Gecatogomphius kavejevi Vjuschkov et Tchudinov, 1957; cf. Riabininus uralensis (Riabinin, 1915). Captorhinomorpha (Bolosaurida), Bolosauridae Cope, 1878: Timanosaurus ivachnenkoi Gubin, 1993. Theromorpha (Dinocephalia), Eotitanosuchidae Tchudinov, 1960: Biarmosuchus sp. Theromorpha (Dinocephalia), Archaeosyodontidae Ivachnenko, fam. nov.: Microsyodon orlovi Ivachnenko, 1995. Theromorpha (Gorgonopia), Phthinosuchidae Efremov, 1954: Kamagorgon ulanovi Tatarinov, 1999. Theromorpha (Gorgonopia), Rhopalodontidae Seeley, 1894: Parabradysaurus udmurticus Efremov, 1954 and P. silantjevi Ivachnenko, 1995. Judging from the biomorph composition, the terrestrial community corresponding to this assemblage was undoubtedly a pioneer coastal community, as in the previous case. The major distinction in the composition of

the terrestrial block is the presence of primitive Gorgonopia (Phthinosuchidae, Rhopalodontidae), as was reconstructed for the initial communities of South Africa. This probably corroborates a very early character of the Mezen Assemblage and the fact that it persisted in refuges. It is possible to propose that the flattened western coast of the East European basin (which preceded the Kazanian Sea) was occupied much earlier than the eastern coast adjacent to the Paleo-Ural Mountains. The eastern slope has yielded a block of ichthyophagous hydrophiles (Melosauridae–Archaeosyodontidae), which formed a special aquatic community. In this case, it is possible to say that the subassemblage belongs to a pioneer megacommunity. As in the previous case, direct analogues of this assemblage in other territories have not been recorded (Ivakhnenko, 2005a). Natural development of the Golyusherma Subassemblage resulted in the formation of the Ocher Faunal Subassemblage, which comprises the following taxa: Batrachomorpha (Edopiformes), Melosauridae Fritsch, 1885: Konzhukovia tarda Gubin, 1991. Batrachomorpha (Edopiformes), Archegosauridae Meyer, 1858: Collidosuchus tchudinovi Gubin, 1986. Batrachomorpha (Edopiformes), Platyoposauridae Huene, 1931: Platyoposaurus stuckenbergi (Trautschold, 1884) and Bashkirosaurus cherdyncevi Gubin, 1981. Batrachomorpha (Dissorophoidea), Dissorophidae Boulenger, 1902: Zygosaurus lucius Eichwald, 1848; Kamacops acervalis Gubin, 1980; and Iratusaurus vorax Gubin, 1980. Parareptilia (Pareiasaurida), Nycteroleteridae Romer, 1956: Bashkyroleter bashkyricus (Efremov, 1940). Parareptilia (Pareiasaurida), Tokosauridae Tverdochlebova et Ivachnenko, 1984: Tokosaurus perforatus Tverdochlebova et Ivachnenko, 1984. Parareptilia (Pareiasaurida), Rhipaeosauridae Tchudinov, 1955: Rhipaeosaurus tricuspidens Efremov, 1940. Anthracosauromorpha (Gephirostegia), Enosuchidae Konzhukova, 1955: Enosuchidae gen. indet. Captorhinomorpha (Bolosaurida), Belebeyidae Ivachnenko, 2001: Belebey vegrandis Ivachnenko, 1973; B. maximi Tverdochlebova, 1987; and Davletkulia gigantea Ivachnenko, 1990. Theromorpha (Dinocephalia), Eotitanosuchidae Tchudinov, 1960: Biarmosuchus tener Tchudinov, 1960 and B. tchudinovi Ivachnenko, 1999. Theromorpha (Dinocephalia), Archaeosyodontidae Ivachnenko, fam. nov.: Archaeosyodon praeventor Tchudinov, 1960. Theromorpha (Gorgonopia), Phthinosuchidae Efremov, 1954: Dinosaurus murchisoni (Fischer, 1845). Theromorpha (Gorgonopia), Rhopalodontidae Seeley, 1894: Rhopalodon wangenheimi Fischer, 1841 and Phthinosaurus borissiaki Efremov, 1940. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

Theromorpha (Gorgonopia), Estemmenosuchidae Tchudinov, 1960: Estemmenosuchus uralensis Tchudinov, 1960 and E. mirabilis Tchudinov, 1968. Theromorpha (Anomodontia), Venyukoviidae Efremov, 1940: Otsheria netzvetajevi Tchudinov, 1960. The differences in taxonomic composition from the previous subassemblage are relatively insignificant; in some cases, they are distinctly connected with a much more thorough examination of localities. The major character is the appearance of Eotitanosuchidae and Estemmenosuchidae. Regarding the general morphological pattern, these groups are direct derivatives or close relatives of the groups recorded earlier in eastern Europe. However, they form a coevolutionary trophic pair of peculiar biomorphs, a giant phytophagous saprophagan and specialized predator, and form the large-sized, dominant, block (Olson, 1952, 1976, 1983; Sennikov, 1995, 1996). The subdominant block is composed of the same biomorphs as primitive pioneer communities. In fact, the two blocks were independent communities, which were not connected trophically. The presence of separate dominant and subdominant communities is characteristic of typical Permian oligobiomorph megacommunities of sea coasts and lowlands, including aquatic communities (Ivakhnenko, 2006). The analysis of the taxonomic composition shows that the Ocher Faunal Assemblage is mostly endemic (Ivakhnenko, 2002c; Kalandadze and Rautian, 2002). As mentioned above, it resembles in the Dinomorpha composition the fauna of South Africa, but includes representatives of special evolutionary trends, i.e., Rubidgeoidea (instead of Gorgonopioidea), Rhopalodontoidea (instead of Burnetioidea), and Ulemicia (instead of Dicynodontia). At the same time, the presence in eastern Europe of derivatives of a number of tetrapod groups (archegosauroid labyrinthodonts, bolosaurids, primitive parareptiles) known in the Early Permian of western Europe but absent from Gondwana suggests that the initial center of the formation of faunas of the eastern coast of the Kazanian basin is a presently unknown region of the northern slope of the Tethys. If this is the case, the initial Gondwanan fauna was probably similar to the fauna of the southern slope of this ocean, but differed somewhat in composition. The next stage in the development of the East European fauna is represented by the Isheevo Faunal Assemblage. It comprises the following taxa: Batrachomorpha (Edopiformes), Melosauridae Fritsch, 1885: Konzhukovia vetusta (Konzhukova, 1955); Tryphosuchus paucidens Konzhukova, 1955; T. kinelensis (Vjuschkov, 1955); and Uralosuchus tverdochlebovae Gubin, 1993. Batrachomorpha (Edopiformes), Platyoposauridae Huene, 1931: Platyoposaurus vjuschkovi Gubin, 1989. Parareptilia (Pareiasaurida), Lanthanosuchidae Efremov, 1946: Lanthanosuchus watsoni Efremov, 1946; Chalcosaurus rossicus Meyer, 1866; and C. lukjanovae (Ivachnenko, 1980). PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

975

Anthracosauromorpha (Gephyrostegia), Enosuchidae Konzhukova, 1955: Enosuchus breviceps Konzhukova, 1955. Captorhinomorpha (Bolosaurida), Belebeyidae Ivachnenko, 2001: Permotriturus herrei Tatarinov, 1968. Theromorpha (Nikkasauria), Microuraniidae Ivachnenko, 1995: Microurania minima Ivachnenko, 1995 and M. mikia Ivachnenko, 2003. Theromorpha (Dinocephalia), Syodontidae Ivachnenko, 1994: Syodon gusevi (Tchudinov, 1968) and S. efremovi (Orlov, 1940). Theromorpha (Dinocephalia), Anteosauridae Boonstra, 1954: Titanophoneus potens Efremov, 1938; T. adamanteus (Orlov, 1958); and T. rugosus (Trautschold, 1884). Theromorpha (Dinocephalia), Deuterosauridae Seeley, 1894: Deuterosaurus biarmicus Eichwald, 1846 and D. jubilaei (Nopcsa, 1928). Theromorpha (Dinocephalia), Ulemosauridae Ivachnenko, 1994: Ulemosaurus svijagensis Riabinin, 1938 and U. gigas (Efremov, 1954). Theromorpha (Anomodontia), Ulemicidae Ivachnenko, 1996: Ulemica invisa (Efremov, 1938) and U. efremovi Ivachnenko, 1995. Theromorpha (Therocephalia), Pristerognathidae Broom, 1908: Porosteognathus efremovi Vjuschkov, 1952. It is evident that the assemblage consists mostly of descendants and derivatives of the groups known in the previous assemblages and includes primitive entomophages (Microuraniidae), which could have occurred in earlier assemblages but have not yet been recorded. The presence of the Lanthanosuchidae, which are closely related to Lanthaniscidae of the Mezen Assemblage, is probably connected with the link between the faunas of two coasts of the Kazanian basin in the southern part and penetration of this peculiar biomorph, which has no analogue on the eastern coast. Thus, only minor changes occurred in the composition of subdominant community. At the same time, the dominant community retains the former biomorphs, but changes considerably in taxonomic composition. This is connected with the replacement of the former trophic pair Eotitanosuchidae–Estemmenosuchidae by a new pair, Anteosauridae–Ulemosauridae. It is hardly possible that this replacement was caused by changes in physiographic conditions resulted from expansion of the territory and transformation of the southern East European region into a coastal lowland that covered both slopes of the Kazanian basin. It seems more likely that the southern part of the basin was dried and, hence, a connection with northern Gondwana was established and provided expansion of new groups of a progressive lineage of Dinocephalia. It is interesting that, despite close affinity of the initial phytophagous groups (Ulemosauridae and Riebeeckosauridae), further fate of these lineages in eastern Europe and South Africa differed considerably.

976

IVAKHNENKO

In the first case, the evolution of specialized saprophagans of this group came to end, and peculiar euryphagous Deuterosauridae were formed. In the second case, a much more diverse set of phytophagous taxa was formed (Tapinocephalidae, Strutiocephalidae). This is probably connected with a wider range of suitable ecotopes in South Africa. The Isheevo Faunal Assemblage corresponds to an oligobiomorph megacommunity of a coastal lowland (Ivakhnenko, 2006). These megacommunities are connected with the ecotope type designated placket (Ivakhnenko, 2001), according to the terminology proposed by Vysotskii (1927). These ecotopes are formed on lowlands with a warm humid climate, unstable hydrological regime, and mostly subaqueous landscapes. Judging from the taxonomic composition, the assemblage considered resembles the faunas of the Eodicynodon and Tapinocephalus zones of South Africa. However, as in the previous case, considerable differences in taxonomic composition are observed. These differences are mostly attributable to the set of derivatives preserved of the groups of the previous assemblages. It is noteworthy that South African faunas include Gorgonopidae and Dicynodontidae (Rubidge, 1995), which are absent from eastern Europe of the same time. The taxonomic composition of the subdominant block of these areas sharply differs, except for the presence of primitive therocephals (Pristerognathidae). However, the material of the only East European representative of the group (Porosteognathus) is very poor and unusual; therefore, it may belong to a separate endemic family (Ivakhnenko, 2001, p. 125). The presence of new groups of Dinocephalia, which are relatively close to South African taxa, concerns only the dominant block. Consequently, no direct exchange between faunas was observed in this case; apparently, there was only a link through an intermediate zone. A true faunal link between South Africa and eastern Europe was only observed at the subsequent developmental stage, which is represented by the Sokolki Faunal Assemblage. In regard to eastern Europe, this was undoubtedly connected with an elevation of some area that transformed into a wide band with uniform physiographic conditions, extending from approximately the meridian of Moscow to the Paleo-Ural Mountains and designated the East European Placket (Ivakhnenko, 2001). Because of a remote position from the seacoast, the climate became more continental. Even at the initial stage of the formation of a new fauna, represented by the Kotelnich Faunal Subassemblage, the taxonomic composition changed considerably, although a number of former taxa were preserved. The subassemblage is composed of the following taxa: Parareptilia (Discosauriscomorpha), Kotlassiidae Romer, 1934: Microphon sp. Parareptilia (Pareiasaurida), Nycteroleteridae Romer, 1956: Emeroleter levis Ivachnenko, 1997.

Parareptilia (Pareiasaurida), Bradysauridae Huene, 1948: Deltavjatia vjatkensis (Hartmann-Weinberg, 1937). Anthracosauromorpha (Chroniosuchia), Chroniosuchidae Vjuschkov, 1957: Suchonica vladimiri Golubev, 1999. Theromorpha (Gorgonopia), Phthinosuchidae Efremov, 1954: Viatkogorgon ivakhnenkoi Tatarinov, 1999. Theromorpha (Anomodontia), Galeopidae Broom, 1912: Suminia getmanovi Ivachnenko, 1994. Theromorpha (Anomodontia), Pristerodontidae Toerien, 1953: Australobarbarus kotelnitshi Kurkin, 2000 and A. platycephalus Kurkin, 2000. Theromorpha (Therocephalia), Scylacosauridae Broom, 1903: Kotelcephalon viatkensis Tatarinov, 1999. Theromorpha (Therocephalia), Chthonosauridae Tatarinov, 1974: Viatkosuchus sumini Tatarinov, 1995. Theromorpha (Therocephalia), Perplexisauridae Tatarinov, 2000: Perplexisaurus foveatus Tatarinov, 1997 and Chlynovia serridentatus Tatarinov, 2000. Theromorpha (Therocephalia), Karenitidae Tatarinov, 1997: Karenites ornamentatus Tatarinov, 1995. Theromorpha (Therocephalia), Nanictidopidae Watson et Romer, 1956: Scalopodontes kotelnichi Tatarinov, 2001. Theromorpha (Therocephalia), Scaloposauridae Haughton, 1924: Scalopodon tenuisfrons Tatarinov, 1999. The taxonomic composition of the assemblage shows that the megacommunity changed essentially. Even in the aquatic community, which remained relatively stable throughout the previous history, archegosauroid labyrinthodonts were replaced by the anthracosauromorphs Chroniosuchidae, and aquatic ichthyophagous Dinocephalia were replaced by aquatic therocephals (Chthonosauridae: Ivakhnenko, 2001, p. 118). The dominant community of large saprophagans and predators trophically connected with them completely disappeared. A peculiar united terrestrial community was formed; it includes mostly the amphibiotic phytophages Bradysauridae and Pristerodontidae, predatory Phthinosuchidae, and various small nanophages, mostly therocephals or Nycteroleteridae preserved from the previous assemblages. The major change is the emergence of Dicynodontia (Galeopidae and Pristerodontidae) in eastern Europe. As indicated above, East European Anomodontia (Ulemicia) should not be regarded as ancestors of this group; in addition, they could have originated only in Gondwana, where this group is widespread in earlier assemblages. In the preceding assemblages of eastern Europe, Dinocephalia were represented by diverse biomorphs that occurred in different communities. Therefore, the disappearance of this group in the Kotelnich Subassemblage was not connected with the extinction of invaders; it is more likely that physiographic conditions changed; such that all Dinocephalia (as a particular morphophysiological group) became extinct. The disPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

appearance of Dinocephalia in South Africa occurred at the boundary of the Tapinocephalus–Pristerognathus zones (Rubidge, 1995). If these processes in both relatively remote territories were connected with a certain global factor, they could develop simultaneously and determine the time of faunal link. If elision of Dinocephalia in eastern Europe resulted from, for example, a local increase in continentality of climate, physiographic barrier could have had selective effect on different groups. If this is the case, the time of contact is assigned tentatively to the Tapinocephalus Zone, when Scaloposauria and Dicynodontia expanded in Gondwana. The faunas of eastern Europe and South Africa were most similar in the time of existence of the Ilinskoe Faunal Subassemblage. This subassemblage consists of the following taxa: Batrachomorpha (Colosteiformes), Dvinosauridae Amalitzky, 1921: Dvinosaurus primus Amalitzky, 1921. Parareptilia (Discosauriscomorpha), Kotlassiidae Romer, 1934: Microphon exiguus Ivachnenko, 1983; M. gracilis Bulanov, 2003; and M. arcanus Bulanov, 2003. Parareptilia (Discosauriscomorpha), Karpinskiosauridae Sushkin, 1925: Karpinskiosaurus ultimus (Tchudinov et Vjuschkov, 1956). Parareptilia (Pareiasaurida), Pareiasauridae Seeley, 1888: Proelginia permiana Hartmann-Weinberg, 1937. Anthracosauromorpha (Chroniosuchia), Chroniosuchidae Vjuschkov, 1957: Chroniosaurus dongusensis Tverdochlebova, 1972 and C. levis Golubev, 1998. Archosauromorpha (Protorosauria), Protorosauridae Huxley, 1871: Eorasaurus olsoni Sennikov, 1997. Theromorpha (Gorgonopia), Ictidorhinidae Broom, 1932: Biarmosuchoides romanovi Tverdochlebova et Ivachnenko, 1994; Ustia atra Ivachnenko, 2003. Theromorpha (Gorgonopia), Burnetiidae Broom, 1923: Proburnetia vjatkensis Tatarinov, 1968 and Niuksenitia sukhonensis Tatarinov, 1977. Theromorpha (Gorgonopia), Gorgonopidae Lydekker, 1890: Sauroctonus progressus (Hartmann-Weinberg, 1938) and Suchogorgon golubevi Tatarinov, 2000. Theromorpha (Anomodontia), Galeopidae Broom, 1912: Suminia cf. S. getmanovi Ivachnenko, 1994. Theromorpha (Anomodontia), Dicynodontidae Owen, 1859: Idelesaurus tataricus Kurkin, 2006. Theromorpha (Therocephalia), Scylacosauridae Broom, 1903: Scylacosuchus orenburgensis Tatarinov, 1968. Cynodontia fam. indet. This list lacks elements inherited from the assemblages earlier than the Sokolki Assemblage; primitive Bradysauridae are replaced by Pareiasauridae; Gorgonopidae, Burnetiidae, and Ictidorhinidae of the Gondwanan origin are present. The aquatic community, which is reconstructed based on the composition of the PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

977

assemblage, is probably inherited and mostly aboriginal (Dvinosauridae, Kotlassiidae, Karpinskiosauridae, Chroniosuchidae). The terrestrial community does not show distinct division into blocks isolated trophically according to great distinctions in size that result in the presence of separate communities. The united terrestrial community is composed of dominant and accessory groups obligatory connected with each other. The community of this type, as the megacommunity which included it, was designated the polybiomorph community (Ivakhnenko, 2005b, p. 469). The Ilinskoe Assemblage was probably dominated by phytophagous Dicynodontidae and predatory Gorgonopidae trophically connected with each other. The aquatic community of this assemblage is endemic, whereas the terrestrial components are represented by the taxa in common with faunas of the Tropidostoma–Cistecephalus zones of South Africa (see Rubidge, 1995). No better markers are available; thus, it is only possible to accept a general conformity of these assemblages. The time of the Tropidostoma Zone in South Africa corresponded to a wide distribution of similar taxa of Dicynodontidae (Aulacocephalodontinae), while in the Cistecephalus Zone, the South African fauna acquired the Rubidgeidae, undoubted descendants of the East European Phthinosuchidae. This was probably the time of direct faunal contact between the two territories. This contact influenced the formation of a subsequent fauna, the Sokolki Faunal Subassemblage, which is composed of the following taxa: Batrachomorpha (Colosteiformes), Dvinosauridae Amalitzky, 1921: Dvinosaurus primus Amalitzky, 1921 and D. campbelli Gubin, 2004. Parareptilia (Discosauriscomorpha), Kotlassiidae Romer, 1934: Kotlassia prima Amalitzky, 1921; Microphon sp. Parareptilia (Discosauriscomorpha), Karpinskiosauridae Sushkin, 1925: Karpinskiosaurus secundus (Amalitzky, 1921) and K. ultimus (Tchudinov et Vjuschkov, 1956). Parareptilia (Discosauriscomorpha), Kotlassiidae Romer, 1934: Kotlassia prima Amalitzky, 1921. Parareptilia (Procolophonida), ?Owenettidae Broom, 1939: “Nyctiphruretus” optabilis Bulanov, 2002. Parareptilia (Pareiasaurida), Elginiidae Cope, 1895: Obirkovia gladiator Bulanov et Jashina, 2005. Parareptilia (Procolophonida), Spondylolestidae Ivachnenko, 1979: Suchonosaurus minimus Tverdochlebova et Ivachnenko, 1994 and Kinelia broomi Bulanov, 2003. Parareptilia (Pareiasaurida), Pareiasauridae Seeley, 1888: Scutosaurus tuberculatus (Amalitzky, 1922); S. karpinskii (Amalitzky, 1922); and S. itilensis Ivachnenko et Lebedev, 1987. Anthracosauromorpha (Chroniosuchia), Chroniosuchidae Vjuschkov, 1957: Chroniosuchus paradoxus

978

IVAKHNENKO

Vjuschkov, 1957; C. licharevi (Riabinin, 1962); and Jarilinus mirabilis (Vjuschkov, 1957). Araeoscelidomorpha (Araeoscelia), Weigeltisauridae Kuhn, 1939: Rautiania alexandri Bulanov et Sennikov, 2006 and R. minichi Bulanov et Sennikov, 2006. Theromorpha (Gorgonopia), Gorgonopidae Lydekker, 1890: Pravoslavlevia parva (Pravoslavlev, 1927). Theromorpha (Gorgonopia), Inostranceviidae Huene, 1948: Inostrancevia alexandri Amalitzky, 1922; I. latifrons Pravoslavlev, 1927; and I. uralensis Tatarinov, 1974. Theromorpha (Gorgonopia), Rubidgeidae Broom, 1938: Leogorgon klimovensis Ivachnenko, 2003. Theromorpha (Anomodontia), Dicynodontidae Owen, 1859: Dicynodon trautscholdi Amalitzky, 1922; D. amalitzkii Sushkin, 1922; Elph borealis Kurkin, 1999; Vivaxosaurus permirus Kalandadze et Kurkin, 2000; Delectosaurus arefjevi Kurkin, 2001; D. berezhanensis Kurkin, 2001; and Interpresosaurus blomi Kurkin, 2001. Theromorpha (Therocephalia), Annatherapsididae Kuhn, 1963: Annatherapsidus petri (Amalitzky, 1922). Theromorpha (Therocephalia), Chthonosauridae Tatarinov, 1974: Chthonosaurus velocidens Vjuschkov, 1955. Theromorpha (Cynodontia), Procynosuchidae Broom, 1937: Nanocynodon seductus Tatarinov, 1968; Uralocynodon tverdokhlebovae Tatarinov, 1987; and Cyrbasidon vladimirense Tatarinov, 2003. Theromorpha (Cynodontia), Dviniidae Sushkin, 1928: Dvinia prima Amalitzky, 1922. The composition of the aquatic ichthyophagous community remains almost the same, the terrestrial community retains a polybiomorph structure. At the same time, an unusual community is formed of large aquatic Pareiasauridae (algophagous Scutosaurus) and large predatory macrophages (Inostranceviidae) trophically connected with each other. In fact, this community shows features of an oligobiomorph association, which is endemic to eastern Europe. It could have developed in connected with the presence in eastern Europe of large inland basins (probably freshwater or with low salinity), which are not characteristic of Karroo (viesses, i.e., very shallow sea, v.s.s., see Ivakhnenko, 2001, p. 9). Thus, at this stage, eastern Europe is distinguished by the presence of a very unusual oligobiomorph community in the polybiomorph megacommunity. This assemblage probably approximately corresponds in time to the Dicynodon Zone of South Africa; however, these faunas undoubtedly lacked a direct link. This is supported by a certain increase in the proportion of taxa endemic to eastern Europe (Annatherapsididae, Dviniidae) and a minor role of Rubidgeidae, which almost completely replaced Gorgonopidae in South Africa of that time. The distinct history of the East European fauna to the terminal Permian is evident from the presence of the

Vyazniki Faunal Assemblage, which has no analogue in other territories. It includes the following taxa: Batrachomorpha (Colosteiformes), Dvinosauridae Amalitzky, 1921: Dvinosaurus egregius Shishkin, 1968; D. purlensis Shishkin, 1968. Parareptilia (Discosauriscomorpha), Karpinskiosauridae Sushkin, 1925: Karpinskiosaurus sp. Parareptilia (Pareiasaurida), Elginiidae Cope, 1895: Obirkovia sp. Anthracosauromorpha (Chroniosuchia), Bystrowianidae Vjuschkov, 1957: Bystrowiana permira Vjuschkov, 1957. Anthracosauromorpha (Chroniosuchia), Chroniosuchidae Vjuschkov, 1957: Uralerpeton tverdochlebovae Golubev, 1998. Archosauromorpha (Thecodontia), Proterosuchidae Huene, 1908: Archosaurus rossicus Tatarinov, 1960. Theromorpha (Anomodontia), Dicynodontidae Owen, 1859: Dicynodontidae gen. indet. Theromorpha (Therocephalia), Annatherapsididae Kuhn, 1963: Annatherapsidus sp. Theromorpha (Therocephalia), Scaloposauridae Haughton, 1924: Malasaurus germanus Tatarinov, 2001. Theromorpha (Therocephalia), Moschowhaitsiidae Tatarinov, 1963: Moschowhaitsia vjuschkovi Tatarinov, 1963. Theromorpha (Therocephalia), Nanictidopidae Watson et Romer, 1956: Hexacynodon purlinensis Tatarinov, 1974. The aquatic ichthyophagous community retains almost the same composition. The amphibiotic groups (phytophagous Elginiidae and predatory Bystrowianidae) are well represented. At the same time, an important event of this stage is the disappearance of oligobiomorph community composed of the pair Scutosaurus– Inostrancevia; this is only attributable to the disappearance of suitable ecotopes, vast and shallow basins (viesses). The preservation of many small ponds and streams combined with the disappearance of large shallow basins may result from an increase in continentality of climate connected with an increase in the distance from the seacoast, which was displaced far north, rather than with an increase in aridity (Ochev and Surkov, 2000). This increase in continentality of climate could have been accompanied by aridity and some cooling during certain periods. Apparently, these processes resulted in the disappearance of Gorgonopidae (Ivakhnenko, 2005b, pp. 438, 473) and penetration into the community of more xerophilous groups, the archosaurian reptiles Proterosuchidae. Xerophiles could have evolved previously on more elevated sites, in plakor ecotopes. The change in conditions in the East European Placket resulted in convergence of conditions in placket and plakor ecotopes and, eventually, in extinction of all Permian families of Dinomorpha (Ivakhnenko, 2005a). PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

CHAPTER 7. PRINCIPLES OF SYSTEMATICS The construction of the system of any taxon is an attempt at putting in order the results of studies, represented in a hierarchical scheme based on certain parameters. The major parameter is the morphological syndrome discussed in relation to its evolution (transformation) in time. The evolution of the morphological syndrome is tightly connected with the evolution of biomorphs, since the morphological structure is influenced considerably by adaptation to particular habitats. The system reflects the history of groups, which develops in space and time; thus, the two parameters should also be taken into account. As a result, a system based on four parameters (morphology, biomorph, time, and range) depends on the extent to which each parameter is understood; as new data on any parameter are acquired, the system should be revised. Since it is hardly possible that absolute data on any parameter will be obtained in the foreseeable future, it is impossible to expect a final variant of the system. The system of a group is influenced not only by the data received, but also by new hypotheses and reconstructions. Almost every new study will introduce some changes; thus, a system that was stable for a long time would indicate only the absence of fundamental studies of this group. For the construction of the hierarchical system of Dinomorpha, the rank of higher taxa is of great significance. Therefore, the main result of the study is the recognition that it is impossible to retain the same taxon for all groups that have achieved the obligatory terrestrial (reptilian) physiological grade. At present, it is evident for me that this level was independently passed by a number of tetrapod lineages (for review of the problem, see Ivakhnenko, 2003c, pp. 348–352; 2006). Each lineage certainly had a certain basal group. Judging from the fact that all groups reaching this level are angustitabulars, with separate exoskeletons of the intercapsule roof and the roof of the occipital ring (the tectum synoticum is topographically under the posterior part of the parietal, and the tectum posterius is overlain by the postparietal); these basal groups probably belonged to Anthracosauromorpha. At present, there are no reliable morphological criteria for the moment of transition to the obligatory terrestrial level; however, it is possible to assume that transitional groups, at least those of the major trunks, differed considerably in many important structural features (Ivakhnenko, 2003c, 2005b). In particular, the basal position in Parareptilia was probably occupied by a synpareial group (synpareialia); this restricted subsequent development to the anapsid cranial design. The initial groups of Diapsida and Theromorpha retained the apopareial structural type of the temporal region (apopareialia); therefore, descendants retained the possibility of the formation of the temporal fenestra in place of the former spiracular fissure, i.e., the parapsid type, with the fenestra located above the caudal process of the postorbital, or the synapsid type, with the fenestra PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

979

under this process (or both). However, primitive diapsids lack a trace of seismosensory grooves, and the cavity of the middle ear of this group is formed differently in the postquadrate position. Theromorphs have a periangular cavity, the presence of which is only possible to explain by the preservation of a bony structure connected with the infradental diverticulum of the mandibular seismosensory line. In addition, to the moment of transition to the reptilian morphophysiological level, all Diapsida had acquired an isolating squamous cover (pholidosic group), which almost completely prevented transpiration through the skin; this resulted in essential changes in physiology. On the contrary, primitive theromorphs probably retained a relatively permeable and glandular skin (pilidosic group). This suggests that the obligatory terrestrial level was reached independently in different tetrapod groups; thus, the retention of a single taxon for all groups achieving the reptilian morphophysiological grade is hardly expedient. Therefore, in the present study, Theromorpha is regarded as a separate class of tetrapods, which is characterized by the following characters: angustitabularia, apopareialia, subapsida+synapsida (syn-subapsida), periangularia, and pilidosica. Theromorpha was regarded as a class in some previous works (see, for example, Brink, 1963, 1986; Ivakhnenko, 1987; Ivakhnenko et al., 1997); however, this was not accompanied by definition or comment. The class is divided into Eutherapsida and Eotherapsida (Ivakhnenko, 2003c) ranked as subclasses. In the subclass Eotherapsida, the division into the infraclasses Sphenacomorpha and Dinomorpha are retained (Ivakhnenko, 2003c). INFRACLASS DINOMORPHA

D e fi n i t i o n. Angustitabularia, apopareialia, synsubapsida, periangularia, pilidosica, paraquadratobasalia, streptostilia. C o m p o s i t i o n. Superorders Nikkasauria, Gorgodontia, and Anomodontia. R e m a r k s. The major change compared to the previous scheme (Ivakhnenko, 2003c) is the establishment of Nikkasauria as a separate taxon of primitive nanophages, which retain a relatively high position of the zygomatic arch, with a well-pronounced subapsid incisure, and, probably, the primary streptostyly with simultaneously functioning quadrate–squamosal and quadrate–articular joints. This is the last taxonomic level, at which it is possible to compare groups based on alternative syndromes, i.e., the principle of presence or absence (see Ivakhnenko, 2005b, p. 447). Up to this level, all syndromes included in the definition are present in the designated or alternative variants (in the latter case, a group belongs to a different taxon). At the lower taxonomic levels, the characters are sometimes transitional or incompletely expressed. This concerns the extent of development of the subapsid incisure and other temporal structures (including the structures determining the

980

IVAKHNENKO

superiotemporalia and inferiotemporalia), the development of canines, and changes in dentition. The secondary loss of streptostyly, which is usual for some groups, looks differently, because a gradual restriction of mobility in the quadrate–squamosal joint lacks functional sense. However, if the initial design was accompanied by interruption of strict connection between the quadrate and pterygoid (or epipterygoid), the restoration of this connection (lock of the QQJ-complex) could not occur instantaneously and, hence, the loss of mobility in the complex developed gradually. Apparently, this problem requires additional investigation. Superorder Nikkasauria

D i a g n o s i s. Zygomatic arch more or less elevated, such that subapsid incisure well expressed as functional structure connected with skull fenestration; however, zygomatic process of jugal bone not participating in bordering fenestra. It is possible that the following primitive structural characters are diagnostic: braincase streptostylic (basipterygoid processes long, forming joint with pterygoid region); quadrate–squamosal and quadrate–articular joints retaining mobility; vomer paired, with row of small teeth along medial suture (these features are not known in Microuraniidae). C o m p o s i t i o n. Order Nikkasaurida. R e m a r k s. I tentatively include in this taxon groups differing in dental structure and geological age, such as Nikkasauridae and Microuraniidae; shared characters are the biomorph and general primitive state. This is connected with a poor understanding of the morphology of Microuraniidae. The establishment of this taxon is actually based on the morphology of Nikkasauridae; perhaps, new data on Microuraniidae will result in a revision of the taxonomic position of this group. The superordinal rank of this taxon is only connected with the necessity to oppose it to Gorgodontia and Anomodontia at the same rank; therefore, Nikkasauridae and Microuraniidae are tentatively united in one order; it is possible that, in the future, these groups will be placed in different orders or other higher taxa. Order Nikkasaurida

D i a g n o s i s. The only order of a superorder. C o m p o s i t i o n. Families Nikkasauridae Ivachnenko, 2000 and Microuraniidae Ivachnenko, 1995, eastern Europe.

eroconule type, having one or two notches. Upper jaw sometimes with increased tooth. Small entomophages. G e n e r i c c o m p o s i t i o n. Nikkasaurus Ivachnenko, 2000 and Reiszia Ivachnenko, 2000 from eastern Europe, Mezen Faunal Assemblage, pioneer community of the western slope of the Kazanian basin. Family Microuraniidae Ivachnenko, 1995

T y p e g e n u s. Microurania Ivachnenko, 1995. D i a g n o s i s. Very small animals. Incisors with transversely widened base and flattened occlusal surface. Third upper and fourth lower teeth increased in size, shaped like circular in cross section grasping canine, with well-developed cingulum; cheek teeth with somewhat widened crowns (lateroconule type), with few bordering lateroconules. Small entomophages, possibly combined with sclerophagy. G e n e r i c c o m p o s i t i o n. Microurania Ivachnenko, 1995, eastern Europe, Isheevo Faunal Assemblage, subdominant community of the initial stage (coastal lowland) of the East European Placket. R e m a r k s. Two species of the genus considered (M. minima Ivachnenko, 1995 and M. mikia Ivachnenko, 2003) differ considerably in the dental structure and, particularly, in the incisors (see Ivachnenko, 2003c, pp. 372, 373). Perhaps, they should be assigned to different genera; however, available material is too fragmentary to perform a detailed comparison. Superorder Gorgodontia

D i a g n o s i s. Zygomatic arch horizontal or somewhat lowered, overlapping subapsid incisure. Function of temporal fenestra performed only by synapsid fenestra (monofenestralia). C o m p o s i t i o n. Orders Dinocephalia and Gorgonopia. R e m a r k s. The initial differentiation of two phylogenetic lineages corresponding to the orders Dinocephalia and Gorgonopia was probably connected with differences in the initial feeding adaptations of primitive taxa. The increased pressure in the incisive region of the initial phyto-omnivores resulted in prevailing development of the relatively anterior portions of muscles (anterosynapsida). On the contrary, in the sister group of carni-omnivores, the early formation of the alternative canine caused the displacement of pressure in the region of the anterior margin of the maxillary combined with the predominant development of the posterior portions of muscles (posterosynapsida).

Family Nikkasauridae Ivachnenko, 2000

T y p e g e n u s. Nikkasaurus Ivachnenko, 2000. D i a g n o s i s. Very small animals; temporal fenestra narrow, vertical, widened somewhat posterodorsally. Incisors of primitive centroconic type, cheek teeth flattened conical; posteriormost jaw teeth of lat-

Order Dinocephalia

D i a g n o s i s. Anterosynapsida, i.e., upper border of temporal fenestra displaced anteriorly, crest on posterior margin of postorbital bone (crista postorbitalis) displaced anteriorly on dorsal surface of skull roof, PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

forming fovea supraorbitalis in postorbital region. Temporal fenestra expanding mostly anterodorsally. C o m p o s i t i o n. Suborders Niaftasuchida and Dinocephalida. R e m a r k s. According to the results received, the suborder Niaftasuchida is most similar in morphology to the most primitive Dinocephalia, probably initial to the other groups. Suborder Niaftasuchida

D i a g n o s i s. Temporal fenestra small, almost isometric. Zygomatic arch narrow, almost horizontal, very slightly lowered. Parietal foramen with narrow scalloped edges. Skull with weak klinorhiny, fovea supraorbitalis occupying less than half of dorsal surface of postorbital bone. Anterior incisors dolabriform, inclined anteriorly, with small thickening in place of cinguloid. Cheek teeth of short-leaf type, with few large lateroconules. QQJ-complex probably streptostylic. C o m p o s i t i o n. Family Niaftasuchidae Ivachnenko, 1990, eastern Europe. Family Niaftasuchidae Ivachnenko, 1990

T y p e g e n u s. Niaftasuchus Ivachnenko, 1990. D i a g n o s i s. Family of a monotypic superfamily, amphibiotic frugivores with cheek teeth forming fence structure, and incisors of tearing off type. G e n e r i c c o m p o s i t i o n. Niaftasuchus Ivachnenko, 1990, eastern Europe, Mezen Faunal Assemblage, pioneer community of the western slope of the Kazanian basin. Suborder Dinocephalida

D i a g n o s i s. Temporal fenestra large, vertically extended. Zygomatic arch widened, thickened, lowered. Parietal foramen positioned on high tubercle. Klinorhiny well pronounced, fovea supraorbitalis occupying more than half of dorsal surface of postorbital bone. Anterior incisors almost vertical, with distinct cinguloids. Cheek teeth with many small lateroconules. C o m p o s i t i o n. Infraorders Dinocephalina and Eotitanosuchina. R e m a r k s. The Dinocephalina formed relatively compact group of taxa, South African and East European representatives are very similar in morphology. The opposition to Eotitanosuchina is connected with the early divergence of this predatory lineage, which was probably formed in eastern Europe in conditions of deficiency in predators. Infraorder Dinocephalina

D i a g n o s i s. Alternative cutting canines absent; if upper and lower canine-shaped teeth present, they circular in cross section, at least lower teeth usually retainPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

981

ing traces of cinguloid structures. Cheek teeth always with more or less developed cinguloids. C o m p o s i t i o n. Superfamilies Tapinocephaloidea Owen, 1876; Deuterosauroidea Seeley, 1894; and Titanosuchoidea Broom, 1903. R e m a r k s. The superfamilies display three basic specialization directions of the group, i.e., large phytophagous saprophagous Tapinocephaloidea, specialized predatory Deuterosauroidea, and aquatic ichthyophagous Titanosuchoidea. Judging from the dental structure, the predatory groups evolved from primitive phytophages; however, the differentiation in morphology was less than that of known Tapinocephaloidea; therefore, they are regarded as taxa of the same rank. S u p e r f a m i l y Tapinocephaloidea Owen, 1876

D i a g n o s i s. Temporal fenestra vertical ovate, relatively small because zygomatic arch expanded considerably. Anterior incisors with well-developed cinguloids, extending for more than half of crown diameter. Only upper jaw occasionally having canine-shaped tooth. Cheek teeth with long necks, of short-leaf type, forming fence structure. C o m p o s i t i o n. Families Ulemosauridae Ivachnenko, 1994 from eastern Europe and Tapinocephalidae Owen, 1876 from South Africa. R e m a r k s. South African Tapinocephalidae show a separate trend in the development of the group, evolving from relatively primitive Riebeeckosaurinae Boonstra, 1952 (the temporal region of the skull is relatively elongated; the quadratum–articulare joint is slightly displaced anteriorly, at most to the line of the midlength of the temporal fenestra) to Tapinocephalinae Owen, 1876 (the skull is relatively high and short; the quadratum–articulare joint is displaced considerably anteriorly) and to Strutiocephalinae Haughton, 1929 (the klinorhiny is particularly well pronounced and the preorbital part is elongated and lowered). The upper teeth are not strengthened to acquire canine shape, in contrast to those of Ulemosauridae. Family Ulemosauridae Ivachnenko, 1994

T y p e g e n u s. Ulemosaurus Riabinin, 1938. D i a g n o s i s. Upper jaw with canine-shaped tooth, exceeding in diameter neighboring teeth. Area of quadratum–articulare joint only slightly displaced anteriorly, at most to line of midlength of temporal fenestra. Amphibiotic saprophagans with compressing incisors. G e n e r i c c o m p o s i t i o n. Ulemosaurus Riabinin, 1938, eastern Europe, Isheevo Faunal Assemblage, oligobiomorph community of the initial stage (coastal lowland) of the East European Placket. R e m a r k s. Based on the material from the Middle Permian of eastern Europe, two species, Ulemosaurus svijagensis Riabinin, 1938, lectotype PIN, no. 2207/2, skull, and U. gigas (Efremov, 1954), lectotype PIN,

982

IVAKHNENKO

no. 1955/5, isolated tooth, have been described. This tooth is the third or fourth precanine (incisor) of a very large animal rather than a canine (Ivakhnenko, 2003c, p. 392, text-figs. 20e–20f); perhaps, it belongs to a separate genus. S u p e r f a m i l y Deuterosauroidea Seeley, 1894

D i a g n o s i s. Temporal fenestra large, expanded posteriorly; posterior to parietal foramen, its posterior border abruptly elevated above level of orbits. Cinguloids of anterior incisors small, much smaller than half of crown width. Upper and lower canine-shaped teeth present, somewhat flattened in cross section, with cutting anterior and posterior borders with many rounded lateroconules. Upper “canine” positioned almost vertically in alveolus, slightly curved. Cheek teeth with very short necks, compressed longitudinally and slightly curved posteriorly, with serrated cutting borders. C o m p o s i t i o n. Families Deuterosauridae Seeley, 1894 from eastern Europe and South Africa and Anteosauridae Boonstra, 1954 from eastern Europe, South Africa, and Central Asia. R e m a r k s. Because of typologically similar prey, predators that are closely related but represent different lineages may be very similar in skull design; therefore, it is often difficult to reveal their true relationships. A thorough comparative morphological study of particular taxa is required to gain an insight into this problem. It is possible that Deuterosauria, like Tapinocephalia considered above, formed different lineages, i.e., East European Deuterosaurus–Titanophoneus, South African Tapinocaninus–Anteosaurus, and, probably, Asian (Sinophoneus). They probably independently evolved from omnivores, which retained morphological similarity to phytophages, to specialized predators. However, it may well be that this route was followed only once, and different regions were inhabited by descendents of the same ancestor. At the same time, the resolution of this question is important for the construction of the system, i.e., the assignment of particular lineages to certain taxa and division into initial and specialized groups, or combination of morphologically similar groups in formal taxa. In this case, available published data on South African taxa are insufficient to make an unequivocal solution; therefore, I follow the simplest way, describing formal taxa. Family Deuterosauridae Seeley, 1894

T y p e g e n u s. Deuterosaurus Eichwald, 1848. D i a g n o s i s. Preorbital region of skull short. Zygomatic arch widened, adjoining external surface of lower jaw. Incisors short, with well-developed cinguloids. Upper “canine” long, slightly flattened, thickened at base. Lower “canine” short, circular in cross section, with narrow line of cinguloid retaining lateroconules. Postcanines with widened rounded crowns

having narrow cinguloids. Terrestrial secondary omnivores (bearlike biomorph type). G e n e r i c c o m p o s i t i o n. Deuterosaurus Eichwald, 1848, eastern Europe, Isheevo Faunal Assemblage, subdominant community of the initial stage (coastal lowland) of the East European Placket; Tapinocaninus Rubidge, 1991, South Africa, fauna of the Eodicynodon Zone. R e m a r k s. Tapinocaninus has well-developed upper and lower canine-shaped teeth, widened crowns of cheek teeth, and incisors with well-developed heels (see Rubidge, 1991) and corresponds to the diagnosis of Deuterosauridae of the system accepted here. Two East European species of the genus Deuterosaurus (D. biarmicus Eichwald, 1848, based on specimen PIN, no. 1954/1, incomplete skull, and D. jubilaei (Nopcsa, 1928), holotype PIN, no. 1954/2, very incomplete skull) differ only in proportions (see Ivakhnenko, 2003c, p. 387, text-figs. 17, 18). The two specimens may belong to the same species; therefore, both were used for reconstruction in the present study (Fig. 62b). Family Anteosauridae Boonstra, 1954

T y p e g e n u s. Anteosaurus Watson, 1921. D i a g n o s i s. Preorbital region of skull elongated. Zygomatic arch narrow, not adjoining laterally external surface of lower jaw. Upper margin of angular bone with thickened ridge for muscle. Incisors elongated, with very weak cinguloids. Upper and lower “canines” slightly flattened, having narrow anterior and posterior cutting borders. Lower “canine” retaining thickening in place of cinguloid. Postcanines flattened, curved posteriorly, with many notches in cutting border (“carnassial”). Hygrophilous predators macrophages. G e n e r i c c o m p o s i t i o n. Titanophoneus Efremov, 1938, eastern Europe, Isheevo Faunal Assemblage, oligobiomorph community of the initial stage (coastal lowland) of the East European Placket; Anteosaurus Watson, 1921 (? = Paranteosaurus Boonstra, 1954), South Africa, fauna of the Tapinocephalus Zone; Sinophoneus Cheng et Ji, 1996 (? = Stenocybus Cheng et Li, 1997), Central Asia, northern China, fauna of the Xidagou Formation. R e m a r k s. Available published data on Paranteosaurus (see Boonstra, 1954) and Stenocybus (see Cheng and Li, 1997) suggest to assign them to the same genus. It is possible that the specimens described are only different age stages of the same taxon. S u p e r f a m i l y Titanosuchoidea Broom, 1903

D i a g n o s i s. Temporal fenestra large, subsquare in shape; posterior to line of parietal foramen, its posterodorsal border lowered. Anterior incisors with narrow cinguloids. Upper and lower canine-shaped teeth present, circular in cross section, without cutting borders. Upper “canine” inclined in alveolus, curved PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

strongly posteriorly. Cheek teeth with widened bulbous or petaloid crowns. C o m p o s i t i o n. Families Archaeosyodontidae Ivachnenko, fam. nov., eastern Europe and South Africa; Titanosuchidae Broom, 1903, South Africa; Syodontidae Ivachnenko, 1994, eastern Europe; and Biseridentidae Li et Cheng, 1997, Central Asia. R e m a r k s. It is even more difficult to reveal true relationships in this group than in Deuterosauroidea. These ichthyophagous predators are very similar in general skull design; at the same time, they are rather diverse. The significance of differences remains uncertain because of undoubtedly insufficient published data. Perhaps, they reflect adaptations of particular taxa and are connected with different evolutionary trends in individual lineages, for example, those of South Africa, eastern Europe, and Central Asia. In eastern Europe, the evolutionary trend Microsyodon–Archaeosyodon– Syodon is clearly seen in both age and morphological changes (see above). In this case, early primitive Archaeosyodontidae are easily opposed to younger groups (see Ivakhnenko, 2003c). Archaeosyodon is very similar to Australosyodon from the Eodicynodon Zone of South Africa (see Rubidge, 1994); therefore, the last genus is included in this taxon. These genera are almost identical in the extent to which muscles penetrated into the parietal region and in the structure of the “canine” and cheek teeth and differ somewhat only in the skull proportions. At the same time, the later South African genus Jonkeria van Hoepen, 1916 from the Tapinocephalus Zone (see Broom, 1929) is very similar in dental structure to Archaeosyodontidae (see Boonstra, 1962); however, it is unique in the structure of the flattened and widened preorbital part of the skull, and its wide parietal region looks much more primitive than that of known Archaeosyodontidae. Biseridens Li et Cheng, 1997 from China is similar to Syodon in the structure of the temporal fenestra, which is elevated in the parietal region, with the abruptly lowered posterodorsal margin; however, its cheek teeth are very unusual (the “canine” structure is not known), and the character of pachyostosis is distinct from other taxa (Li and Cheng, 1997). These contradictory data complicate the development of hypotheses concerning the phylogeny of this group; therefore, I accept a formal division of the superfamily into a number of families.

983

G e n e r i c c o m p o s i t i o n. Microsyodon Ivachnenko, 1995, eastern Europe, Golyusherma Faunal Subassemblage, aquatic community of the eastern slope of the Kazanian basin; Archaeosyodon Tchudinov, 1960, eastern Europe, Ocher Faunal Subassemblage, aquatic community of the eastern slope of the Kazanian basin; and Australosyodon Rubidge, 1994, South Africa, fauna of the Eodicynodon Zone. Family Syodontidae Ivachnenko, 1994

T y p e g e n u s. Syodon Kutorga, 1838. D i a g n o s i s. Upper border of temporal fenestra coming onto tubercle of parietal foramen, crista postorbitalis of postorbital bone coming onto surface of postfrontal. Upper “canine” cylindrical, strongly curved. Cheek teeth having thickened crowns and lacking pronounced necks. Anterior edge of skull not widened, pachyostosis of cranial bones very weak. Ichthyophagous hydrobionts of otterlike biomorph type, with pressing cheek teeth. G e n e r i c c o m p o s i t i o n. Syodon Kutorga, 1838, eastern Europe, Isheevo Faunal Assemblage, aquatic community of the initial stage (coastal lowland) of the East European Placket. R e m a r k s. The type species of the genus Syodon, S. biarmicum Kutorga, 1838, is only represented by the anterior part of the skull; “Notosyodon” gusevi Tchudinov, 1968 is represented by the occipital region, and both taxa differ from S. efremovi (Orlov, 1940), in which a complete skull is known. Therefore, the three taxa were tentatively assignment to one genus (see Ivakhnenko, 2003c, pp. 380, 381, text-figs. 9–11), while the generic composition of the family remains an open question. Infraorder Eotitanosuchina

D i a g n o s i s. Alternative upper and lower canines present, teardrop-shaped in cross section, cutting, equally developed. Cheek teeth without cinguloids. C o m p o s i t i o n. Families Eotitanosuchidae Tchudinov, 1960 and Alrausuchidae Ivachnenko, fam. nov., eastern Europe. Family Eotitanosuchidae Tchudinov, 1960

Family Archaeosyodontidae Ivachnenko, fam. nov.

T y p e g e n u s. Archaeosyodon Tchudinov, 1960. D i a g n o s i s. Upper border of temporal fenestra terminating short of parietal foramen, crista postorbitalis of postorbital bone occupying entire dorsal surface of postorbital. Upper “canine” conical, relatively short and slightly curved. Cheek teeth with long necks. Anterior edge of skull not widened, pachyostosis of cranial bones not developed. Ichthyophagous hydrobionts of otterlike biomorph type, with grasping cheek teeth. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

T y p e g e n u s. Biarmosuchus Tchudinov, 1960. D i a g n o s i s. Canines very long relative to skull size, slightly curved. QQJ-complex streptostylic, with articular block between capitulum quadrati and fossa quadratica of squamosal; quadratum–articulare joint immobile. Auditory apparatus of bicrural–quadrate design. Hygrophilous predators macrophages. G e n e r i c c o m p o s i t i o n. Biarmosuchus Tchudinov, 1960, eastern Europe, pioneer community of the Golyusherma Faunal Subassemblage and oligobio-

984

IVAKHNENKO

morph community of the Ocher Faunal Subassemblage of the eastern slope of the Kazanian basin. R e m a r k s. The designation of the type specimen of B. tener Tchudinov, 1960 and, hence, the generic name (Eotitanosuchus or Biarmosuchus: Tchudinov, 1960) was considered in the previous study by Ivakhnenko (1999, 2003c, p. 374). B. tchudinovi Ivachnenko, 1999 is only represented by the maxillary bones; however, the structure of long, slightly curved canine and narrow postcanines formally determine the assignment to this family. Perhaps, new finds will result in the establishment of a separate genus or even family, since the relatively small size of this animal is in contrast with the biomorph of the family considered. Family Alrausuchidae Ivachnenko, fam. nov.

T y p e g e n u s. Alrausuchus Ivachnenko, gen. nov. D i a g n o s i s. Canines relatively short, strongly curved. QQJ-complex nonstreptostylic, quadratum– articulare joint mobile only at lateral condyle of quadrate bone. Auditory apparatus of bicrural–postquadrate design. Medium-sized terrestrial predators–generalists of doglike biomorph type. G e n e r i c c o m p o s i t i o n. Alrausuchus Ivachnenko, gen. nov., eastern Europe, Mezen Faunal Assemblage, pioneer community of the western slope of the Kazanian basin. R e m a r k s. The only known species of this genus was originally described as Biarmosuchus tagax Ivachnenko, 1990, holotype PIN, no. 3706/10 (Peza-1 locality, Arkhangelsk Region, Russian Federation; Middle Permian, Urzhumian Stage). The above mentioned reexamination resulted in the assignment of this species to a separate genus, the diagnosis of which is the same as that of the monotypic family. Order Gorgonopia

D i a g n o s i s. Posterosynapsida, i.e., marginal crest of postorbital (crista postorbitalis) passing along its caudal margin, sometimes displaced even onto its medial surface. Temporal fenestra expanding mostly posteriorly. C o m p o s i t i o n. Suborders Ictidorhinida, Gorgonopida, and Estemmenosuchida. Suborder Ictidorhinida

D i a g n o s i s. Temporal fenestra elongated ovate, vertical; its upper border positioned lower than frontoparietal region. Zygomatic arch short, slightly lowered. Skull nonklinorhinal. Streptostyly of QQJ-complex possibly retained. Alternative canines poorly developed, flattened; postcanines of long-leaf type. C o m p o s i t i o n. Family Ictidorhinidae Broom, 1932, South Africa and eastern Europe.

Family Ictidorhinidae Broom, 1932

T y p e g e n u s. Ictidorhinus Broom, 1913. D i a g n o s i s. Family of monotypic suborder. Small terrestrial carnivorous predators–generalists of martenlike biomorph type. G e n e r i c c o m p o s i t i o n. Ictidorhinus Broom, 1913, South Africa, fauna of the Dicynodon Zone; Rubidgina Broom, 1942, South Africa, fauna of the Cistecephalus and Dicynodon zones; Ustia Ivachnenko, 2003; and Biarmosuchoides Tverdochlebova et Ivachnenko, 1994, eastern Europe, Ilinskoe Faunal Subassemblage, terrestrial community of the higher stage (placket plain) of the East European Placket. R e m a r k s. Perhaps, the group represents late relicts of representatives of Gorgonopia that are very primitive in morphology and followed a long peculiar evolutionary pathway. This conclusion follows from the combination of the primitive structure of the temporal region and primitive teeth with the presence of the interparietal strengthening the sagittal suture. Suborder Gorgonopida

D i a g n o s i s. Temporal fenestra posteriorly widened, its upper border not lower than frontoparietal region. Zygomatic arch long, slightly lowered. Skull nonklinorhinal. Streptostyly of QQJ-complex retained. Alternative canines well developed, flattened, cutting. Postcanines flattened conical, cutting, curved posteriorly, with many cutting notches (“carnassial”). C o m p o s i t i o n. Superfamilies Rubidgeoidea Broom, 1938 and Gorgonopioidea Lydekker, 1890. S u p e r f a m i l y Rubidgeoidea Broom, 1938

D i a g n o s i s. Interparietal bone absent (or very small, irregular in shape). Postorbital arch expanded, zygomatic arch lowered and abruptly widened, with lateral curvature in squamosal part, so that plane of temporal fenestra oriented obliquely upwards. C o m p o s i t i o n. Families Phthinosuchidae Efremov, 1954; Inostranceviidae Huene, 1948, eastern Europe; and Rubidgeidae Broom, 1938, eastern Europe, South and East Africa. Family Phthinosuchidae Efremov, 1954

T y p e g e n u s. Dinosaurus Fischer, 1847. D i a g n o s i s. Skull moderately elongated, lengthto-width ratio at least 1.3 : 1. Orbits very large, their diameter at least one-fifth of skull length. Postorbital arch slightly widened, at most one-third of orbital diameter. Palatal tubercles broad, with many small teeth. At least seven postcanines present. Mediumsized and large terrestrial predators–generalists. G e n e r i c c o m p o s i t i o n. Kamagorgon Tatarinov, 1999, eastern Europe, Golyusherma Faunal Subassemblage, subdominant community of the eastern PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

slope of the Kazanian basin; Dinosaurus Fischer, 1847, eastern Europe, Ocher Faunal Subassemblage, subdominant community of the eastern slope of the Kazanian basin; Admetophoneus Efremov, 1954, eastern Europe, Isheevo Faunal Assemblage, subdominant community of the initial stage (coastal lowland) of the East European Placket; and Viatkogorgon Tatarinov, 1999, eastern Europe, Kotelnich Faunal Subassemblage, terrestrial community of the higher stage (placket plain) of the East European Placket. Family Rubidgeidae Broom, 1938

D i a g n o s i s. Skull relatively short, length-towidth ratio at most 1.2 : 1. Orbits relatively small, much smaller in diameter than one-sixth of skull length. Postorbital arch strongly expanded, at least half of orbital diameter. Palatal tubercles (particularly pterygoid tubercle) narrow, teeth reduced considerably. At most six postcanines present. C o m p o s i t i o n. Subfamilies Rubidgeinae Broom, 1938 from South Africa and eastern Europe and Broomicephalinae Tatarinov, 1974 from South Africa. R e m a r k s. Broomicephalinae (Broomicephalus Brink et Kitching, 1953) is distinguished by the very short skull, expanded in the occipital region; the skull length is less than its greatest width. Subfamily Rubidgeinae Broom, 1938

T y p e g e n u s. Rubidgea Broom, 1938. D i a g n o s i s. Skull moderately widened in temporal region, skull length always greater than its width. Large terrestrial predators–generalists of tigerlike biomorph type. G e n e r i c c o m p o s i t i o n. Sycosaurus Haughton, 1924; Leontocephalus Broom, 1940, South and East Africa, fauna of the Dicynodon Zone; Dinogorgon Broom, 1936, South Africa, fauna of the Cistecephalus and Dicynodon zones, East Africa, fauna of the Dicynodon Zone; Rubidgea Broom, 1938; Prorubidgea Broom, 1840; Clelandina Broom, 1948, South Africa, fauna of the Cistecephalus and Dicynodon zones; Cephalocustriodus Parrington, 1974, East Africa, fauna of the Cistecephalus Zone; and Leogorgon Ivachnenko, 2003, eastern Europe, Sokolki Faunal Subassemblage, terrestrial community of the higher stage (placket plain) of the East European Placket. R e m a r k s. The assignment of Leogorgon to this family (Ivakhnenko, 2003c, p. 406) is tentative, because available material is fragmentary (holotype PIN, no. 4549/13, braincase). Perhaps, this is late large East European derivative of Phthinosuchidae, which evolved independently of South African Rubidgeidae. Family Inostranceviidae Huene, 1948

T y p e g e n u s. Inostrancevia Amalitzky, 1922. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

985

D i a g n o s i s. Skull narrow elongated in projection, length-to-width ratio 2 : 1, correlating with very long temporal fenestra, with posteriorly extended posterodorsal corners. Orbits medium-sized, with diameter one-sixth of skull length. Postorbital arch moderately widened, at most one-third of orbital diameter. Palatal tubercles absent, palatal teeth reduced. Upper jaw with at most four postcanines, lower jaw lacking postcanines. Hygrophilous predators–macrophages. G e n e r i c c o m p o s i t i o n. Inostrancevia Amalitzky, 1922, eastern Europe, Sokolki Faunal Subassemblage, oligobiomorph community of the higher stage (placket plain) of the East European Placket. R e m a r k s. Inostrancevia uralensis Tatarinov, 1974 represented by a very fragmentary material (braincases) was assigned to the genus Inostrancevia based on the presence of a narrow elongated paroccipital process, which corresponds to other species of this genus and is connected with the broad parietal region. S u p e r f a m i l y Gorgonopioidea Lydekker, 1890

D i a g n o s i s. Interparietal bone large, rhombic. Postorbital arch not widened; zygomatic arch narrow, curved gently downwards. Plane of temporal fenestra almost lateral. C o m p o s i t i o n. Families Galesuchidae Watson et Romer, 1956; Cyonosauridae Tatarinov, 1974, South Africa; and Gorgonopidae Lydekker, 1890, South and East Africa, eastern Europe. R e m a r k s. The foundation of the division of the group into three families and diagnoses were given in the previous study by Ivakhnenko (2005b). Family Gorgonopidae Lydekker, 1890

D i a g n o s i s. Preorbital part of skull moderately elongated, ratio of length of preorbital part to width at canines ranging from 1.6 to 2. Interorbital width approximately one-third of skull roof length (from anterior margin to occipital edge). C o m p o s i t i o n. Subfamilies Gorgonopinae Lydekker, 1890 and Scylacopinae Watson et Romer, 1956, South and East Africa, eastern Europe. R e m a r k s. The establishment of the subfamilies was validated in the previous study (Ivakhnenko, 2005b). The subfamilies differ only in skull proportions. The establishment of taxa within the subfamilies is even more difficult, because the group comprises forms having much in common and differing in mosaic of characters of uncertain taxonomic value. Apparently, this is primarily connected with a great number of similar microbiomorphs formed by gorgonopians in the Late Permian, when gorgonopids were most abundant and almost the only group of medium-sized predators. Most likely, there were many minor structural variations; however, against a background of well-developed morphological design of a universal predator–general-

986

IVAKHNENKO

ist, particular distinctions are very difficult to reveal in isolated skulls, which are usually distorted and poorly preserved. Therefore, only the most typical and relatively thoroughly investigated forms should be considered within the composition of the subfamilies. Subfamily Gorgonopinae Lydekker, 1890

T y p e g e n u s. Gorgonops Owen, 1876. D i a g n o s i s. Orbits relatively small, interorbital width at least one-third of skull roof length (from anterior margin to occipital edge); longitudinal orbital diameter at most one-third as long as preorbital part. Medium-sized terrestrial predators–generalists of doglike biomorph type. G e n e r i c c o m p o s i t i o n. Gorgonops Owen, 1876, South Africa, fauna of the Tapinocephalus– Cistecephalus zones; Aelurognathus Haughton, 1924, South Africa, fauna of the Cistecephalus Zone; Arctops Watson, 1914; Arctognathus Broom, 1911, South Africa, fauna of the Cistecephalus Zone; and Pravoslavlevia Vjuschkov, 1953, eastern Europe, Sokolki Faunal Subassemblage, terrestrial community of the higher stage (placket plain) of the East European Placket. Subfamily Scylacopinae Watson et Romer, 1956

T y p e g e n u s. Scylacops Broom, 1913. D i a g n o s i s. Orbits relatively large, interorbital width at most one-third of skull roof length (from anterior margin to occipital edge); longitudinal orbital diameter at least one-third as long as preorbital part. Medium-sized terrestrial predators–generalists of doglike biomorph type. G e n e r i c c o m p o s i t i o n. Lycaenops Broom, 1925, South and East Africa, fauna of the Tropidostoma–Dicynodon zones; Scylacops Broom, 1913; Aloposaurus Broom, 1910, South Africa, fauna of the Cistecephalus Zone; Sauroctonus Bystrov, 1955; and Suchogorgon Tatarinov, 1999, eastern Europe, Ilinskoe Faunal Subassemblage, terrestrial community of the higher stage (placket plain) of the East European Placket. Suborder Estemmenosuchida

D i a g n o s i s. Temporal fenestra widened posteriorly, although its upper margin positioned lower than frontoparietal plane. Zygomatic arch long, strongly lowered. Klinorhiny well-pronounced (basipalatal angle between palatal plane and braincase bottom less than 180°). QQJ-complex nonstreptostylic. Alternative canines weakly developed, occasionally shaped like tusks circular in cross section. Postcanines more or less petaloid, with a few projections–denticles. C o m p o s i t i o n. Superfamilies Burnetioidea Broom, 1923 and Rhopalodontoidea Seeley, 1894.

R e m a r k s. The two superfamilies correspond to two evolutionary trends, the South African Burnetioidea and East European Rhopalodontoidea, the nearest common ancestors of which belong to primitive groups, probably closely related to Hipposauridae and Rhopalodontidae. In eastern Europe, Rhopalodontidae appeared very early in the paleontological record and, losing the cutting borders of the canine and increasing in size, evolved into specialized phytophagous Estemmenosuchidae (Ocher Faunal Assemblage). The evolutionary peak of the South African trend was relatively small omnivores Burnetiidae; this group appeared in eastern Europe much later, in the early phases of the Sokolki Faunal Assemblage. S u p e r f a m i l y Burnetioidea Broom, 1923

D i a g n o s i s. Klinorhiny relatively weak, lower cheek teeth only slightly inclined. Canines flattened, with serrated borders. C o m p o s i t i o n. Families Hipposauridae Watson et Romer, 1956, South Africa, and Burnetiidae Broom, 1923, South Africa and eastern Europe. Family Burnetiidae Broom, 1923

T y p e g e n u s. Burnetia Broom, 1923. D i a g n o s i s. Bones of skull roof with well-developed pachyostotic thickenings shaped like crests and tubercles in frontonasal and supraorbital regions and on squamosal and zygomatic bones. Pachyostotic thickenings particularly well-developed on posterior margin of postorbital arch and upper border of temporal fenestra. Terrestrial omnivores (possibly, with carnivorous specialization of piglike biomorph type). Generic c o m p o s i t i o n. Styracocephalus Haughton, 1929; Bullacephalus Rubidge et Kitching, 2003, South Africa, fauna of the Tapinocephalus Zone; Lobalopex Sidor, Hopson et Keyser, 2004, South Africa, fauna of the Tropidostoma Zone; Lemurosaurus Broom, 1949, South Africa, fauna of the Cistecephalus Zone; Burnetia Broom, 1923, South Africa, fauna of the Dicynodon Zone; Proburnetia Tatarinov, 1968; and Niuksenitia Tatarinov, 1977, eastern Europe, Ilinskoe Faunal Subassemblage, terrestrial community of the higher stage (placket plain) of the East European Placket. S u p e r f a m i l y Rhopalodontoidea Seeley, 1894

D i a g n o s i s. Klinorhiny well pronounced, lower cheek teeth considerably inclined posteriorly. Canines without serrated borders, faceted or circular in cross section. C o m p o s i t i o n. Families Rhopalodontidae Seeley, 1894 and Estemmenosuchidae Tchudinov, 1960, eastern Europe. Family Rhopalodontidae Seeley, 1894

T y p e g e n u s. Rhopalodon Fischer, 1841. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

D i a g n o s i s. Canine longitudinally faceted, bordering crests preserved, although borders lacking serration. Postcanines with high petaloid crowns of longleaf type. Teeth relatively large, dentary at most three times as high as middle postcanines. Labial area on dentary narrow, postcanine row displaced lingually at most to line of internal edge of canine base. Terrestrial secondary frugivores. G e n e r i c c o m p o s i t i o n. Parabradysaurus Efremov, 1954, eastern Europe, Golyusherma Faunal Subassemblage, pioneer community of the eastern slope of the Kazanian basin; Rhopalodon Fischer, 1841; Phthinosaurus Efremov, 1940, eastern Europe, Ocher Faunal Subassemblage, subdominant community of the eastern slope of the Kazanian basin. R e m a r k s. Unfortunately, Rhopalodontidae are only represented in East European collections by isolated jaw bones; therefore, the diagnoses for this group are based almost exclusively on dental characters. Judging from a great number of isolated jaw fragments, this group was probably widespread in respective communities. Family Estemmenosuchidae Tchudinov, 1960

T y p e g e n u s. Estemmenosuchus Tchudinov, 1960. D i a g n o s i s. Canines circular in cross section, without trace of cutting borders or edges. Postcanines with low petaloid crowns, of short-leaf type. Teeth relatively small, dentary more than five times as high as postcanines. Labial area on margin of dentary wide, postcanine row displaced far inside from internal edge of canine base, passing anteriorly lingual to alveolus of canine. Amphibiotic saprophagans with grasping incisor apparatus. G e n e r i c c o m p o s i t i o n. Estemmenosuchus Tchudinov, 1960, eastern Europe, Ocher Faunal Subassemblage, oligobiomorph community of the eastern slope of the Kazanian basin. R e m a r k s. In addition to the type species Estemmenosuchus uralensis Tchudinov, 1960, the same locality has yielded E. mirabilis Tchudinov, 1968. The type species is represented by very many isolated bones and skulls of animals of different individual age, whereas E. mirabilis is only represented by the holotype (PIN, no. 1758/6, incomplete skeleton). This suggests that the locality is close to autochthonous with reference to Estemmenosuchus uralensis, in contrast to E. mirabilis; thus, these taxa may belong to essentially different biomorphs. Morphological distinctions between the two species are rather great. The skull of E. mirabilis is shorter and wider, and the zygomatic arch deviates somewhat laterally, providing passage onto the dentary surface for an external portion of muscles that reaches a special oblique rugose crest on the upper margin of the angular bone. It also considerably differs in the pattern of pachyostotic horns on the skull roof bones. These profound morphological differences suggest that PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

987

the two taxa may belong to different genera; however, the presence of only two forms complicates the establishment of a new genus. Superorder Anomodontia

D i a g n o s i s. Zygomatic arch curving dorsally, subapsid incisure large, zygomatic process of jugal participating in bordering this incisure. Synapsid fenestra (upper) and subapsid incisure (lower) playing role of temporal fenestra (difenestralia). C o m p o s i t i o n. Orders Ulemicia and Dicynodontia. R e m a r k s. The taxonomic scheme developed in the present study is based on the revision of King (1988), with some modifications. East European Ulemicia (= Venjukovioidea Watson et Romer, 1956) is regarded as a separate order opposed to the South African lineage (Dicynodontia); according to the scheme accepted here, they are ranked orders. Order Ulemicia

D i a g n o s i s. Articular surface of condyle of articular bone concave, restricting jaw movements to compression. C o m p o s i t i o n. Families Venyukoviidae Efremov, 1940 and Ulemicidae Ivachnenko, 1996. Family Venyukoviidae Efremov, 1940

T y p e g e n u s. Venyukovia Amalitzky, 1922. D i a g n o s i s. Premaxillary region of palate high, vaulted. Tooth plates of maxillary bones narrow. Anteriorly, bony choanae widely open, ovate. Jaw teeth high, narrow cylindrical. Posterior margin of subapsid incisure not extending into area above quadrate. Terrestrial specialized frugivores with shearing jaw system and cutting self-sharpening incisors. G e n e r i c c o m p o s i t i o n. Otsheria Tchudinov, 1960 and Venyukovia Amalitzky, 1922, eastern Europe, Ocher Faunal Subassemblage, subdominant community of the eastern slope of the Kazanian basin. R e m a r k s. Perhaps, the two genera of the family are synonyms; however, the type species are represented by hardly comparable specimens and, hence, it seems inexpedient to combine them in the same genus. Family Ulemicidae Ivachnenko, 1996

T y p e g e n u s. Ulemica Ivachnenko, 1996. D i a g n o s i s. Premaxillary region of palate low. Dental plates of maxillary bones wide. Anteriorly, bony choanae abruptly narrowed by wide maxillary plates. Jaw teeth shortly conical, low. Posterior border of subapsid incisure coming into area above quadrate. Terrestrial specialized frugivores, with compressing jaw system and cutting self-sharpening incisors.

988

IVAKHNENKO

G e n e r i c c o m p o s i t i o n. Ulemica Ivachnenko, 1996, eastern Europe, Isheevo Faunal Assemblage, subdominant community of the initial stage (coastal lowland) of the East European Placket. Order Dicynodontia

D i a g n o s i s. Articular surface of condyle of articular bone convex, providing sliding longitudinal movements of jaws. C o m p o s i t i o n. Suborders Dromasaurida and Dicynodontida. Suborder Dromasaurida

D i a g n o s i s. Crowns of jaw teeth large relative to jaw bones (at least one-third as high as dentary); teeth completely developed in both jaws and in premaxillaries, participating in food treatment. C o m p o s i t i o n. Family Galeopidae Broom, 1912, South Africa and eastern Europe. R e m a r k s. Apparently, Patranomodon Rubidge et Hopson, 1990, with relatively small teeth, elongated skull, and less elevated zygomatic arch, should be assigned to a separate family. The possibility of longitudinal jaw movements (see Reisz and Sues, 2000) suggests to assign it to Dromasaurida; however, this genus requires additional study. Family Galeopidae Broom, 1912

T y p e g e n u s. Galeops Broom, 1912. D i a g n o s i s. Skull short, zygomatic arch elevated above lower orbital border. Teeth large, pointed petaloid, with serrated borders. Premaxillary incisors enlarged, directed obliquely anteriorly. Terrestrial specialized frugivores (colonial subfossorial?). G e n e r i c c o m p o s i t i o n. Galeops Broom, 1912, South Africa, fauna of the Tapinocephalus Zone; Anomocephalus Modesto, Rubidge et Welman, 1999, South Africa, fauna of the Eodicynodon Zone; Suminia Ivachnenko, 1994, eastern Europe, Kotelnich and Ilinskoe faunal subassemblages, terrestrial community of the higher stage (placket plain) of the East European Placket.

sible to recognize two basic patterns of jaw apparatus, i.e., Diictodon-like (mostly shearing) and Dicynodonlike (mostly grinding). This conclusion mostly agrees with the data obtained by Cluver and King (1983, textfig. 40) and King (1988, p. 70), who regarded Endothiodon as the initial morphotype that gave rise to two basic lineages, i.e., groups of taxa related to Diictodon and to Dicynodon. These results suggest the presence of three basic groups in the suborder considered. The infraorder Endothiodontina is relatively close to the initial morphotype (retaining almost complete tooth row in the maxillary and dentary), while two major advanced lineages show considerable reduction (down to complete disappearance) of cheek teeth. The lineages Diictodontina and Dicynodontina probably differ in the major direction of specialization of the jaw apparatus (see above). Endothiodon and related taxa (see King, 1988, pp. 79–81) should not be taken for the initial group for the others because they are late representatives that retain a number of primitive features, in particular, relatively well-developed jaw teeth. The considerable independent evolution of the group is evident from an increase in size of the temporal cavity, with the formation of a narrow parietal crest and reduced pterygoid flanges. These features are expressed to a greater extent than in primitive representatives of two other superfamilies. Representatives of Endothiodontina have not been recorded in eastern Europe; the same is true of representatives of Diictodontina (including the families Emydopidae van Hoepen, 1934, Diictodontidae Cluver et King, 1983, Cistecephalidae Seeley, 1895, Myosauridae Cluver et King, 1983, etc.: see King, 1988, pp. 114–124). I am not inclined to place Eodicynodon Barry, 1974 in a special group of the same rank as the above taxa. The characters of this form (low jaw arch, paired vomers, etc.: see Barry, 1974; Cluver and King, 1983) show that it is merely primitive, according to its early geological age (fauna of the Eodicynodon Zone). It is possible that the genus in question is only a primitive representative of one of two later lineages (most likely Dicynodontina). Infraorder Dicynodontina

Suborder Dicynodontida

D i a g n o s i s. Jaw teeth reduced; if retained, their crowns many times lower than dentary height; at least in premaxillary part, teeth replaced by horn sheath of jaws. C o m p o s i t i o n. Infraorders Endothiodontina, Diictodontina, and Dicynodontina. R e m a r k s. The taxonomic system of Dicynodontida is insufficiently developed. Apparently, it is expedient to construct the system based on detailed analysis of work of the jaw apparatus, which substantially determines the general skull design. Surkov (2005) has shown that, based on the functioning of jaws, it is pos-

D i a g n o s i s. Dorsal surface of dentaries with deep groove. Palatines considerably contributing to construction of premaxillary–maxillary grinding region (Dicynodon-like grinding dentition). C o m p o s i t i o n. Superfamilies Dicynodontoidea Owen, 1859; Lystrosauroidea Cope, 1870; and Kannemeyerioidea Huene, 1948. R e m a r k s. In the present study, three major stages of improvement of the general skull design are recognized and tentatively ranked as superfamilies: Dicynodontoidea, primitive Permian taxa with very poorly pronounced klinorhiny; Lystrosauroidea, taxa from the PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

Permian–Triassic boundary with a very well-pronounced klinorhiny; and Kannemeyerioidea, Triassic taxa with intermediate development of klinorhiny. The second and third (Triassic) superfamilies are not considered in the present study. S u p e r f a m i l y Dicynodontoidea Owen, 1859

D i a g n o s i s. Skull relatively long, temporal region at least half as long as skull. Klinorhiny poorly pronounced, angle between braincase bottom and palatal plane at least 170°. C o m p o s i t i o n. Families Pristerodontidae Toerien, 1953 and Dicynodontidae Owen, 1859. R e m a r k s. The general evolutionary trend of this group was gradual reduction of cheek teeth up to complete replacement by palatal cornification and, hence, an increase in relative area of the palatine plate of the premaxillary bone and strengthening of adductors, which occupied a more vertical position, increasing the pressure, while grinding reciprocal jaw movements with optimum amplitude were retained. It is possible to divide this process into stages, which are regarded as taxa of family rank. It is probable that, at each evolutionary stage, morphological changes developed in similar directions; this is evident from undoubtedly independent increase in weight of adductors through narrowing the parietal region up to the formation of a relatively narrow parietal crest. Family Pristerodontidae Toerien, 1953

T y p e g e n u s. Pristerodon Huxley, 1868. D i a g n o s i s. Cheek teeth retained, functioning as fence structure, bordering laterally mouth cavity. Anterior edge of bony choana approximately in line with midlength of maxillary. Primitive terrestrial herbivores. G e n e r i c c o m p o s i t i o n. Pristerodon Huxley, 1868, South Africa, fauna of the Tapinocephalus–Dicynodon zones; Emydurans Broom, 1921, South Africa, fauna of the Tropidostoma Zone; Australobarbarus Kurkin, 2000, eastern Europe, Kotelnich Faunal Subassemblage, terrestrial community of the higher stage (placket plain) of the East European Placket. Family Dicynodontidae Owen, 1859

D i a g n o s i s. Cheek teeth completely reduced or retained as isolated rudiments lacking functional significance. Anterior edge of bony choana in line with posterior margin of maxillary or even posterior to it. Terrestrial efficient herbivores. C o m p o s i t i o n. Subfamilies Aulacocephalodontinae Toerien, 1953, South Africa and eastern Europe, and Dicynodontinae Owen, 1859, East and South Africa, eastern Europe, Central Asia. R e m a r k s. At present, it is almost impossible to develop a satisfactory system of the family in connecPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

989

tion with the early formation in the group of a general optimum skull design of the efficient phytophage–herbivore, with a grinding system of horny jaw sheath. The group forms very many biomorphs and microbiomorphs, which correspond to either particular specializations to certain plant types or realization of similar but somewhat different improvements of the jaw apparatus for the treatment of coarse fibrous plants. This corresponds to a wide range of morphologically similar designs, distinguished mostly by the proportions of the skull as a whole and its segments. In addition, it is extremely difficult to estimate the taxonomic significance of particular characters and, the more so, to gain an insight into the hierarchy of syndromes. The construction of the system is particularly complicated because of the absence of easily interpretable cranial structures, which were directly connected with the mode of life (such as teeth in phytophagous mammals). It is only possible to construct the system of the group based on a thorough study of a representative series of skulls; however, at present, this is extremely difficult. Apparently, technical estimates of general skull designs will be of great importance; even taxa very similar externally differ somewhat in the proportions of the main cranial segments. I believe that calculations of this kind will provide invaluable data on the distribution of stresses, differences in these parameters, and, hence, on the functioning of the jaw apparatus. Certainly, the differences in work of the jaw apparatus caused eventually all seemingly insignificant differences in design and proportions. The differences in the ratio of the skull length and its width in the temporal region are most prominent; this parameter divides the group into taxa with wide or narrow skulls and suggests to regard them as separate subfamilies. The subfamilies distinguished by such an unstable and weakly reliable character are considered to include only most typical and relatively well understood taxa, which are characterized below. Subfamily Aulacocephalodontinae Toerien, 1953

T y p e g e n u s. Aulacocephalodon Seeley, 1898. D i a g n o s i s. Skull expanded considerably laterally in temporal region, so that greatest skull width equal to, or even greater than, length. Temporal fenestra rounded, almost isometric. G e n e r i c c o m p o s i t i o n. Aulacocephalodon Seeley, 1898, South Africa, fauna of the Cistecephalus– Dicynodon zones; Pelanomodon Broom, 1938, South Africa, fauna of the Dicynodon Zone; Tropidostoma Broom, 1915, South Africa, fauna of the Tropidostoma Zone; and Idelesaurus Kurkin, 2006, eastern Europe, Ilinskoe Faunal Subassemblage, terrestrial community of the higher stage (placket plain) of the East European Placket. Subfamily Dicynodontinae Owen, 1859

T y p e g e n u s. Dicynodon Owen, 1844.

990

IVAKHNENKO

VFA

R3 V2

SFS O2

G5

G4

G3

F5

IFS

V1

G2

KFS N2

I2 F4

R2

IFA

S2

O1

S1 F3

OFS

R1

F2 I1

GFS MFA

F1

G1

N1

Fig. 69. Phylogram of morphological and biomorph evolution of East European Dinomorpha. Designations: (MFA) Mezen Faunal Assemblage, (GFS) Golyusherma Faunal Subassemblage, (OFS) Ocher Faunal Subassemblage, (IFA) Isheevo Faunal Assemblage, (KFS) Kotelnich Faunal Subassemblage, (IFS) Ilinskoe Faunal Subassemblage, (SFS) Sokolki Faunal Subassemblage, and (VFA) Vyazniki Faunal Assemblage. Biomorphs and their representatives: (F) frugivores: (F1) Niaftasuchidae, (F2) Rhopalodontidae, (F3) Venyukoviidae, (F4) Ulemicidae, and (F5) Galeopidae; (G) predators–generalists: (G1) Alrausuchidae, (G2) Phthinosuchidae, (G3) Rubidgeidae, (G4) Ictidorhinidae, and (G5) Gorgonopidae; (I) ichthyophagists: (I1) Archaeosyodontidae and (I2) Syodontidae; (N) nanophages: (N1) Nikkasauridae and (N2) Microuraniidae; (O) omnivores: (O1) Deuterosauridae and (O2) Burnetiidae; (R) predators–macrophages (R1) Eotitanosuchidae, (R2) Anteosauridae, and (R3) Inostranceviidae; (S) saprophagans: (S1) Estemmenosuchidae and (S2) Ulemosauridae; and (V) herbivores (V1) Pristerodontidae and (V2) Dicynodontidae. Arrows designate Gondwanan invaders.

D i a g n o s i s. Skull narrow in temporal region, its greatest width much less than length. Temporal fenestra elongated ovate. G e n e r i c c o m p o s i t i o n. Dicynodon Owen, 1845, East and South Africa, fauna of the Dicynodon Zone; Central Asia, fauna of the Jijeao Formation; western Europe, fauna of the Elgin Formation; eastern Europe, Sokolki Faunal Subassemblage and Vyazniki Faunal Assemblage, terrestrial community of the higher stage (placket plain) of the East European Placket; Oudenodon Owen, 1860, South Africa, fauna of the Cistecephalus–Dicynodon zones; Rhachiocephalus Owen, 1876, South Africa, fauna of the Tropidostoma–Cistecephalus zones; Dinanomodon Broom, 1938, South Africa, fauna of the Cistecephalus–Dicynodon zones; Vivaxosaurus Kalandadze et Kurkin, 2000, Delectosaurus Kurkin, 2001, and Elph Kurkin, 1999, Sokolki Faunal Subassemblage, terrestrial community of the higher stage (placket plain) of the East European Placket.

R e m a r k s. The validity of African and Asian genera is accepted based on the data provided by King (1988) and Lucas (2001). The distribution of South African taxa is given after Rubidge (1995). CONCLUSIONS Figure 69 shows a scheme of relationships between morphological and biomorph evolution of East European Dinomorpha in view of distribution of taxa in the sequence of faunal assemblages. In regard to this topic, family-rank taxa seem most informative. In this scheme, if there is no direct or indirect evidence of adventive nature of a family, it is regarded as native. Only groups that are diverse in earlier faunas of South Africa are considered to be invaders. However, some other groups may also be an adventive result of differentiation in close areas, which presently remain unknown (Ivakhnenko, 2005a). PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA)

It is evident from the scheme that the evolution of Dinomorpha is divided into several stages. The initial stages were passed prior to the appearance of the group in eastern Europe. The stages are as follows: (I) the establishment of the group at morphological and biomorph levels that are close to the later nanophages Nikkasauria; (II) early divergence into Anomodontia and Gorgodontia; (III) division into primarily phytophagous Dinocephalia and primarily predatory Gorgonopia. At the end of stage III, probably already in eastern Europe, Eotitanosuchina, a native lineage of predatory Dinocephalia, was formed. Some primitive groups of these stages (Nikkasauridae, Niaftasuchidae, Alrausuchidae) are retained in the earliest fauna, i.e., in the Mezen Faunal Assemblage of the western coast of the Kazanian basin. The Golyusherma Faunal Subassemblage of the eastern coast mostly corresponds to the next stage (IV), the time of the formation of a native phytophagous lineage of primitive Gorgonopia (Rhopalodontidae) and hydrophilous ichthyophagous Dinocephalia (Archaeosyodontidae). At this stage, Anomodontia were probably divided into Ulemicia and Dicynodontia. The Ocher Faunal Subassemblage displays a special stage (V) in the evolution of Dinomorpha, which, according to available data, was observed only in eastern Europe. In this assemblage, the derivatives of primitive phytophages Rhopalodontidae and predatory Alrausuchidae form an oligobiomorph community composed of the trophic pair Estemmenosuchidae–Eotitanosuchidae. It is believed that, in Gondwana of approximately the same time, Dicynodontida evolved from primitive Dromasaurida, while Gorgonopioidea and Burnetioidea evolved from groups closely related to Ictidorhinida. Stage VI is marked by a rapid differentiation of Dinocephalina into phytophagous Tapinocephaloidea and predatory Deuterosauroidea. It is evident that the region of the formation of these groups was uniform for East European and South African representatives; however, considerable morphological differences between East European Ulemosauridae and South African Tapinocephalidae suggest that this region was beyond both known ranges (Ivakhnenko, 2005a). At the same time, the presence of primitive Dinocephalina in early faunas of eastern Europe (Niaftasuchidae) suggest that this process could have been connected with eastern Europe. Here, derivatives of Tapinocephaloidea and Deuterosauroidea formed a new oligobiomorph community of the trophic pair Ulemosauridae–Anteosauridae (Isheevo Faunal Assemblage). Apparently, at this stage in Gondwana, Dicynodontida was divided into a number of lineages (Endothiodontina, Diictodontina, Dicynodontina), with adaptive radiation within them. The next stage (VII) is characterized primarily by complete extinction of Dinocephalia (which occurred almost simultaneously in both areas) rather than by the appearance of particular new groups. This disaster had very significant consequences for eastern Europe. In the Kotelnich and Ilinskoe faunal subassemblages, native elements are only represented by primitive predPALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

991

ators–generalists Phthinosuchidae, while all other Dinomorpha groups are undoubtedly Gondwanan invaders (Ictidorhinidae, Gorgonopidae, Burnetiidae, Galeopidae, Pristerodontidae, and Aulacocephalodontinae). This was probably the time of the formation of Rubidgeidae, large relatives of Phthinosuchidae, which migrated to Gondwana. The last evolutionary stage (VIII) in the history of East European Dinomorpha was marked by the appearance of large macrophagous predators Inostranceviidae in the Sokolki Faunal Subassemblage, which formed an endemic relict oligobiomorph community (Scutosaurus–Inostrancevia). The final stage (IX) was characterized by almost complete extinction of Paleozoic groups at the Permian–Triassic boundary. Of the Paleozoic families, only Emydopidae passed into the Early Triassic of Gondwana (Myosaurus: Rubidge, 1995). In eastern Europe, this stage is recorded even in the terminal Vyazniki Faunal Assemblage, where only the family Dicynodontidae is retained. ACKNOWLEDGMENTS This study was supported by the Russian Foundation for Basic Research (project no. 07-04-00907), the Board of the President of the Russian Federation for Support of Leading Scientific Schools (project no. NSh-6228.2006.4), and the programs of the Presidium of the Russian Academy of Sciences “Biodiversity” and “Origin and Evolution of Biosphere.” REFERENCES 1. E. F. Allin, “Evolution of the Mammalian Middle Ear,” J. Morphol. 147, 403–437 (1975). 2. E. F. Allin, “The Auditory Apparatus of Advanced Mammal-like Reptiles and Early Mammals,” in The Ecology and Biology of Mammal-like Reptiles, Ed. by N. Hotton, III, P.D. McLean, J.J. Roth, and C.E. Roth (Smithsonian Inst., Washington, 1986), pp. 283–294. 3. V. P. Amalitzky, “Diagnoses of the New Forms of Vertebrates and Plants from the Upper Permian of North Dvina,” Izv. Ross. Akad. Nauk 25 (1), 1–12 (1922). 4. R. T. Bakker, “Juvenile–Adult Habitat Shift in Permian Fossil Reptiles and Amphibians,” Science 217, 53–55 (1982). 5. H. R. Barghusen, “The Origin of the Mammalian Jaw Apparatus,” in Morphology of the Maxillo–Mandibular Apparatus, Ed. by G.H. Schumacher (VEB G. Thieme, Leipzig, 1972), pp. 26–32. 6. H. R. Barghusen, “The Adductor Jaw Musculature of Dimetrodon (Reptilia, Pelycosauria),” J. Paleontol. 47 (5), 823–834 (1973). 7. T. H. Barry, “A New Dicynodont Ancestor from the Upper Ecca (Lower Middle Permian) of South Africa,” Ann. S. Afr. Mus. 74, 117–136 (1974). 8. L. D. Boonstra, “Some Features of the Cranial Morphology of the Tapinocephalid Deinocephalians,” Bull. Am. Mus. Natur. Hist. 72 (2), 75–98 (1936). 9. L. D. Boonstra, “Paranteosaurus gen. nov.; a Titanosuchian Reptile,” Ann. S. Afr. Mus. 42, 157–159 (1954).

992

IVAKHNENKO

10. L. D. Boonstra, “The Dentition of the Titanosuchian Dinocephalians,” Ann. S. Afr. Mus. 46 (6), 57–112 (1962). 11. L. D. Boonstra, “Early Dichotomes in the Therapsida,” Ann. S. Afr. Mus. 59 (5), 176–195 (1963a). 12. L. D. Boonstra, “Diversity within South African Dinocephalia,” S. Afr. Mus. 59 (5), 196–206 (1963b). 13. A. S. Brink, “Speculations on Some Advanced Mammalian Characteristics in the Higher Mammal-like Reptiles,” Palaeontol. Afr. 4, 77–85 (1956). 14. A. S. Brink, “The Taxonomic Position of the Synapsida,” S. Afr. J. Sci. 59 (5), 153–159 (1963). 15. A. S. Brink, “Illustrated Bibliographical Catalogue of the Synapsida,” Repts. S. Afr. Geol. Surv. 10 (1), 1–150 (1986). 16. R. L. Broom, “On the Use of the Term Anomodontia,” Rec. Albany Mus. 1, 266–268 (1905). 17. R. L. Broom, “On a New Cynodont Reptile (Aelurosuchus browni),” Trans. S. Afr. Phil. Soc. 16, 376–378 (1906). 18. R. L. Broom, “On Some New Fossil Reptiles from the Karroo Beds of Victoria West, South Africa,” Trans. S. Afr. Phil. Soc. 18 (1), 31–42 (1907). 19. R. L. Broom, “On the Carnivorous Mammal-like Reptile of the Family Titanosuchidae,” Ann. Transvaal Mus. 13, 9–36 (1929). 20. B. A. Bykov, Ecological Dictionary (Nauka, Alma-Ata, 1983) [in Russian]. 21. Zh. Cheng and J. Li, “A New Genus of Primitive Dinocephalian: the Third Report on Late Permian Dashankou Lower Tetrapod Fauna,” Vertebr. Palasiat. 35 (1), 35–43 (1997). 22. P. K. Chudinov, “New Facts about the Fauna of the Upper Permian of the USSR,” J. Geol. Chicago 73 (1), 117–130 (1965). 23. N. M. Chumakov, “Climate and Climatic Zones of the Permian and Early Triassic,” Tr. Geol. Inst. Ross. Akad. Nauk 550 (Climate in the Epochs of Major Biosphere Transformation), 230–256 (2004). 24. M. A. Cluver, “The Cranial Morphology of the Lower Triassic Dicynodont Myosaurus gracilis,” Ann. S. Afr. Mus. 66 (4), 35–64 (1974). 25. M. A. Cluver and G. M. King, “A Reassessment of the Relationships of Permian Dicynodontia (Reptilia, Therapsida) and a New Classification of Dicynodonts,” Ann. S. Afr. Mus. 91 (3), 195–273 (1983). 26. E. H. Colbert, “The Mammal-like Reptile Lycaenops,” Bull. Am. Mus. Natur. Hist. 89 (6), 357–404 (1948). 27. E. D. Cope, “The Theromorphous Reptilia,” Am. Natur. 12, 829–830 (1878). 28. R. B. Cowles, “Possible Origin of Dermal Temperature Regulation,” Evolution 12, 347–357 (1957). 29. C. B. Cox, “On the Anatomy of a New Dicynodont Genus with Evidence of the Position of the Tympanum,” Proc. Zool. Soc. London, Ser. B 132 (3), 321–367 (1959). 30. C. B. Cox, “A New Digging Dicynodont from the Upper Permian of Tanzania,” in Studies in Vertebrate Evolution, Ed. by K.A. Joysey and T.S. Kemp (Oliver and Boyd, Edinburg, 1972), pp. 173–189.

31. A. W. Crompton, “The Cranial Morphology of a New Genus and Species of Ictidosaurian,” Proc. Zool. Soc. London, Ser. B 130 (2), 183–216 (1958). 32. A. W. Crompton and N. Hotton, III, “Functional Anatomy of Masticatory Apparatus of Two Dicynodonts (Reptilia, Therapsida),” Postilla Peabody Mus. Natur. Hist. 109, 1–51 (1967). 33. A. W. Crompton and F. A. Jenkins, “Origin of Mammals,” in Mesozoic Mammals: The First Two-Thirds of Mammalian History, Ed. by J.A. Lillegraven, Z. KielanJaworowska, and W.A. Clemens (Univ. California Press, Berkley, 1979), pp. 59–73. 34. A. W. Crompton, C. R. Taylor, and J. A. Jagger, “Evolution of Homeothermy in Mammals,” Nature 272 (5651), 333–336 (1978). 35. G. R. De Beer, “The Development of the Vertebrate Skull,” (Oxford Univ. Press, London, 1937). 36. R. De Mar and H. R. Barghusen, “Mechanics and the Evolution of the Synapsid Jaw,” Evolution 26, 622–637 (1972). 37. I. A. Efremov, “Some New Permian Reptiles from the USSR,” Dokl. Akad. Nauk SSSR 19 (9), 771–776 (1938). 38. I. A. Efremov, “New Finds of Permian Terrestrial Vertebrates in Bashkortostan and the Chkalovsk Region,” Dokl. Akad. Nauk SSSR 27 (4), 412–415 (1940a). 39. I. A. Efremov, “Dinocephalian Fauna of the Village of Isheevo,” Tr. Paleontol. Inst. Akad. Nauk SSSR 10 (Preliminary Descriptions of New Forms of the Permian and Triassic Fauna of Terrestrial Vertebrates of the USSR), 31–73 (1940b). 40. I. A. Efremov, “A Terrestrial Vertebrate Fauna from the Permian Copper Sandstones of the Western ForeUrals,” Tr. Paleontol. Inst. Akad. Nauk SSSR 54, 1–416 (1954). 41. E. D. Eichwald, “Über die Saurier des Kupfers ührenden Zechsteins Russlands,” Bull. Soc. Imp. Natur. Moscou 21 (3), 141–150 (1848). 42. E. D. Eichwald, Lethaea Rossica ou Paleontologie de la Russie, Vol. 1: Ancienne periode (Stuttgart, 1860). 43. G. Fischer von Waldheim, “Beitrag zur näheren Bestimmung des von Hrn. Wangenheim von Qualen abgebildeten und besreibenen Saurier-Kopf,” Bull. Soc. Imper. Natur. Moscou 18 (3), 540–543 (1845). 44. G. Fischer von Waldheim, “Bemerkungen über das Schädel-fragment, welches Herr Major Wang. v. Qualen in dem West-Ural entdeckt und der Gesellschaft zur Beurteiling vorgelegt hat,” Bull. Soc. Imper. Natur. Moscou 20 (3), 263–267 (1847). 45. E. Gaupp, “Die Entwicklung des Kopfskeletts,” Hetwig’s Handbuch des Entw.-Lehre 3, 573–875 (1906). 46. V. K. Golubev, “Permian Tetrapod Stratigraphy,” New Mexico Mus. Natur. Hist. Sci. Bull. 30, 299–301 (2005). 47. A. P. Hartmann-Weinberg, “Gorgonopids from the USSR As the Time Markers,” in Problems of Paleontology (Mosk. Gos. Univ., Moscow, 1938), Vol. 4, pp. 47–87 [in Russian]. 48. J. A. Hopson, “The Origins of the Mammalian Middle Ear,” Am. Zool. 6, 437–450 (1966). 49. N. Hotton, III, “An Alternative to Dinosaur Endothermy: The Happy Wanderer,” in A Cold Look at the Warm-blooded Dinosaurs: Selected Symposia Series 28, PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA) Ed. by R.D.K. Thomas and E.C. Olson (Am. Assoc. Advanc. Sci., 1980), pp. 311–350. 50. N. Hotton, III, E. C. Olson, and R. Beerbower, “Amniote Origin and Discovery of Herbivory,” in Amniote Origins: Completing the Transition to Land, Ed. by S.S. Sumida and K.L. Martin (Academic, New York, 1977), pp. 207–264. 51. F. von Huene, “Short Review of the Lower Tetrapods,” Spec. Publ. R. Soc. S. Afr., Vol. R. Broom Commem., 65–106 (1948). 52. F. von Huene, Paläontologie und Phylogenie der niederen Tetrapoden (Fischer, Jena, 1956), pp. 1–716. 53. M. F. Ivakhnenko, “The Problem of Homology of Some Bones of the Skull Roof of Lower Tetrapods,” Zh. Obshch. Biol. 45 (1), 143–151 (1984). 54. M. F. Ivakhnenko, “Permian Parareptiles of the USSR,” Tr. Paleontol. Inst. Akad. Nauk SSSR 223, 1–159 (1987). 55. M. F. Ivakhnenko, “Principles of the Classification of the Lower Tetrapods,” in Abstracts of the First World Congress of Herpetology (Univ. Kent Press, Canterbury, 1989), p. 150. 56. M. F. Ivakhnenko, “Late Paleozoic Faunal Assemblage of Tetrapods from Deposits of the Basin of the Mezen River,” Paleontol. Zh., No. 4, 81–90 (1990). 57. M. F. Ivakhnenko, “New Late Permian Dromasaurians (Anomodontia) from Eastern Europe,” Paleontol. Zh., No. 1, 77–84 (1994). 58. M. F. Ivakhnenko, “Primitive Late Permian Dinocephalian Titanosuchids of Eastern Europe,” Paleontol. Zh., No. 3, 98–105 (1995a). 59. M. F. Ivakhnenko, “New Primitive Therapsida from the Permian of Eastern Europe,” Paleontol. Zh., No. 4, 110– 119 (1995b). 60. M. F. Ivakhnenko, “Primitive Anomodonts, Venyukovias, from the Late Permian of Eastern Europe,” Paleontol. Zh., No. 4, 77–84 (1996). 61. M. F. Ivakhnenko, “Biarmosuches from the Ocher Faunal Assemblage of Eastern Europe,” Paleontol. Zh., No. 3, 79–86 (1999) [Paleontol. J. 33 (3), 309–316 (1999)]. 62. M. F. Ivakhnenko “Estemmenosuches and Primitive Theriodonts from the Late Permian,” Paleontol. Zh., No. 2, 79–86 (2000a) [Paleontol. J. 34 (2), 189–197 (2000)]. 63. M. F. Ivakhnenko, “The Nikkasauridae—Problematic Primitive Therapsids from the Late Permian of the Mezen Localities,” Paleontol. J. 34 (Suppl. 2), 179–186 (2000b). 64. M. F. Ivakhnenko, “Tetrapods from the East European Placket—Late Paleozoic Natural Territorial Complex,” Tr. Paleontol. Inst. Ross. Akad. Nauk 283, 1–200 (2001). 65. M. F. Ivakhnenko, “The Origin and Early Divergence of Therapsids,” Paleontol. Zh, No. 2, 49–57 (2002a) [Paleontol. J. 36 (2), 168–175 (2002)]. 66. M. F. Ivakhnenko, “Taxonomy of East European Gorgonopians (Therapsida),” Paleontol. Zh., No. 3, 56–65 (2002b) [Paleontol. J. 36 (3), 283–292 (2002)]. 67. M. F. Ivakhnenko, “Eotherapsids in the Zoocenoses of the Ocher Faunal Assemblage,” in IV All-Russia Conference on Paleontology and Stratigraphy of the Permian and Triassic of Northern Eurasia, Ed. by L.P. Tatarinov PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

993

and V.K. Golubev (Paleontol. Inst. Ross. Akad. Nauk, Moscow, 2002c), pp. 16–17 [in Russian]. 68. M. F. Ivakhnenko, “The Features of Lower Jaw Articulation in the Gorgonopian Suchogorgon (Therapsida),” Paleontol. Zh., No. 1, 48–52 (2003a) [Paleontol. J. 37 (1), 50–54 (2003)]. 69. M. F. Ivakhnenko, “The Angular Region of the Lower Jaw in Primitive Therapsids,” Paleontol. Zh., No. 3, 65–70 (2003b) [Paleontol. J. 37 (3), 287–292 (2003)]. 70. M. F. Ivakhnenko, “Eotherapsids from the East European Placket,” Paleontol. J. 37 (Suppl. 4), 339–465 (2003c). 71. M. F. Ivakhnenko, “Comparative Survey of Lower Permian Tetrapod Faunas of Eastern Europe and South Africa,” No. 1, 70–75 (2005a) [Paleontol. J. 39 (1), 66–71 (2005)]. 72. M. F. Ivakhnenko, “Morphology of the Gorgonopidae (Eotherapsida) and Tetrapod Communities in the Late Paleozoic,” Paleontol. J. 39 (Suppl. 4), 393–511 (2005b). 73. M. F. Ivakhnenko, “Evolution of Late Paleozoic Tetrapods As Evolution of Their Biomorphs,” in Evolution of the Biosphere and Biodiversity, Ed. by S.V. Rozhnov (KMK, Moscow, 2006), pp. 373–393 [in Russian]. 74. M. F. Ivakhnenko, V. K. Golubev, Yu. M. Gubin, et al., “Permian and Triassic Tetrapods of Eastern Europe,” Tr. Paleontol. Inst. Ross. Akad. Nauk 268, 1–215 (1997). 75. M. F. Ivakhnenko and G. I. Tverdokhlebova, “A Revision of Permian Bolosauromorphs from Eastern Europe,” Paleontol. Zh., No. 2, 98–106 (1987). 76. N. N. Kalandadze and A. A. Kurkin, “A New Permian Dicynodont and the Question of the Origin of the Kannemeyeroidea,” Paleontol. Zh., No. 6, 60–68 (2000) [Paleontol. J. 34 (6), 642–649 (2000)]. 77. N. N. Kalandadze and A. S. Rautian, “Intercontinental and Transtemporal Faunal Relationships of the Ocher and Mezen Faunal Assemblages,” in IV All-Russia Conference on Paleontology and Stratigraphy of the Permian and Triassic of Northern Eurasia, Ed. by L.P. Tatarinov and V.K. Golubev (Paleontol. Inst. Ross. Akad. Nauk, Moscow, 2002), pp. 18–20 [in Russian]. 78. T. S. Kemp, “On the Functional Morphology of the Gorgonopsid Skull,” Phil. Trans. R. Soc. London, Ser. B 256 (801), 1–83 (1969). 79. T. S. Kemp, “The Jaw Articulation and Musculature of the Whaitsiid Therocephalia,” in Studies in Vertebrate Evolution, Ed. by K.A. Joysey and T.S. Kemp (Oliver and Boyd, Edinburg, 1972), pp. 213–230. 80. T. S. Kemp “Mammal-like Reptiles and the Origin of the Mammals,” (Academic, London, 1982). 81. T. S. Kemp, “Interrelationships of the Synapsida,” in The Phylogeny and Classification of the Tetrapoda, Vol. 2: Mammals, Ed. by M. J. Benton (Systematic Assoc., 1988), Spec. Vol. 35B, pp. 1–22. 82. K. A. Kermack and F. Mussett, “The Ear in the Mammal-like Reptiles and Early Mammals,” Acta Paleontol. Polon. 23 (1–2), 147–158 (1983). 83. G. M. King, “Dinocephalia, Anomodontia,” in Encyclopedia of Paleoherpetology, Ed. by P. Wellnhofer (Fischer, Stuttgart, 1988), Part 17, pp. 1–175. 84. A. A. Kurkin, “On the Dental Morphology of Deuterosaurus biarmicus Eichw. (Dinocephalia, Synapsida),”

994

IVAKHNENKO

Paleontol. Zh., No. 1, 92–94 (1997) [Paleontol. J. 31 (1), 90–92 (1997)]. 85. A. A. Kurkin, “A New Dicynodont from Excavations on the Malaya Northern Dvina River,” Paleontol. Zh., No. 3, 87–92 (1999) [Paleontol. J. 33 (3), 326–331 (1999)]. 86. A. A. Kurkin, “New Dicynodonts from the Upper Permian of the Vyatka Basin,” Paleontol. J. 34 (Suppl. 2), 203–210 (2000). 87. A. A. Kurkin, “New Late Paleozoic Dicynodonts from the Vyazniki Assemblage of Terrestrial Tetrapods of Eastern Europe,” Paleontol. Zh., No. 1, 53–60 (2001) [Paleontol. J. 35 (1), 53–59 (2001)]. 88. A. A. Kurkin, “A New Dicynodont (Anomodontia, Eotherapsida) from the Upper Permian of Tatarian,” Paleontol. Zh., No. 4, 81–84 (2006) [Paleontol. J. 35 (4), 434–437 (2006)]. 89. S. S. Kutorga, “Beitrag zur Kenntnis der organischen Überreste des Kupfersandsteins am Westlischen Abhänge des Urals,” (St. Petersburg, 1838). 90. J.-L. Li and Zh. Cheng, “First Discovery of Eotitanosuchian (Therapsida, Synapsida) of China,” Vertebr. Palasiat. 35 (4), 268–282 (1997). 91. S.G. Lucas, “Chinese Fossil Vertebrates,” (Columbia Univ. Press, New York, 2001). 92. Zh. Luo and A. W. Crompton, “Transformation of the Quadrate (Incus) Groove the Transition from Nonmammalian Cynodonts to Mammals,” J. Vertebr. Paleontol. 14 (3), 341–374 (1994). 93. S. P. Modesto, “The Skull of the Herbivorous Synapsid Edaphosaurus boanerges from the Lower Permian of Texas,” Palaeontology 38 (1), 213–239 (1995). 94. F. von Nopcsa, “The Genera of Reptiles,” Palaeobiologica 1, 163–188 (1928). 95. V. G. Ochev and M. V. Surkov, “On the Biogeography of Continental Tetrapods on the Eve of the Terminal Permian Extinction,” Tr. Nauch.-Issled. Inst. Geol. Saratov. Gos. Univ., Nov. Ser. 6 71–80 (2000). 96. E. C. Olson, “Origin of Mammals Based upon Cranial Morphology of the Therapsid Suborders,” Geol. Soc. Am. Spec. Pap., No. 55, 1–136 (1944). 97. E. C. Olson, “The Evolution of a Permian Vertebrate Chronofauna,” Evolution 6 (2), 181–196 (1952) 98. E. C. Olson, “Late Permian Terrestrial Vertebrates USA and USSR,” Trans. Am. Phil. Soc. Philadelphia 52 (Part 2, No. 5), 1–224 (1962). 99. E. C. Olson, “The Exploitation of Land by Early Tetrapods,” in Morphology and Biology of Reptiles, Ed. by A.A. Bellairs and C.B. Cox (Academic, London–New York, 1976), pp. 1–30. 100. E. C. Olson, “Biological and Physical Facts in the Dispersal of Permo–Carboniferous Terrestrial Vertebrates,” in Historical Biogeography, Plate Tectonics, and the Changing Environment, Ed. by J. Gray and A.J. Boucot (State Univ. Press, Oregon, 1979), pp. 227–238. 101. E. C. Olson, “Coevolution or Coadaptation? Permocarboniferous Vertebrate Chronofauna,” in Coevolution, Ed. by M.H. Nitecki (Academic, London–New York, 1983), pp. 307–338. 102. Yu. A. Orlov, “Predatory Dinocephalians from the Isheevo Fauna (Titanosuchians),” Tr. Paleontol. Inst. Akad. Nauk SSSR 72, 1–114 (1958).

103. H. F. Osborn, “The Reptilian Subclasses Diapsida and Synapsida and the Early History of the Diaptosauria,” Mem. Am. Mus. Natur. Hist. 1, 449–507 (1903). 104. F. R. Parrington, “Remarks on a Theory of the Evolution of the Tetrapods Middle Ear,” J. Laringol. Otol. 63, 580–595 (1949). 105. F. R. Parrington, “On the Cranial Anatomy of Some Gorgonopsids and the Synapsid Middle Ear,” Proc. Zool. Soc. London, Ser. B 125 (1), 1–40 (1955). 106. F. R. Parrington, “The Origin of Mammals,” Adv. Sci. 24, 165–173 (1967). 107. P. A. Pravoslavlev, “Gorgonopsidae from the Northern Dvina Excavations of V.P. Amalitzky,” in Northern Dvina Excavations of Prof. V.P. Amalitzky, Ed. by A.P. Karpinskii (Akad. Nauk SSSR, Leningrad, 1927a), Vol. 3, pp. 1–118 [in Russian]. 108. P. A. Pravoslavlev, “Skull Roof of Inostrancevia sp.,” Ezheg. Russ. Paleontol. O–va 6, 51–56 (1927b). 109. R. R. Reisz, “Pelycosauria,” in Encyclopedia of Paleoherpetology, Ed. by P. Wellnhofer (Fischer, Stuttgart– New York, 1986), Part 17A, pp. 1–102. 110. R. R. Reisz and H. D. Sues, “Herbivory in Late Paleozoic and Triassic Terrestrial Vertebrates,” in Evolution of Herbivory in Terrestrial Vertebrates: Perspectives from the Fossil Record, Ed. by H.D. Sues (Cambridge Univ. Press, Cambridge, 2000), pp. 9–42. 111. A. N. Riabinin, “Vertebrate Fauna from Upper Permian Deposits of the Basin of the Sviyaga River: 1. A New Dinocephalian Ulemosaurus svijagensis n. gen., n. sp.,” Ezheg. Tsentr. Nauch.-Issled. Muz. Chernyshev 1, 4–40 (1938). 112. A. S. Romer, “Osteology of the Reptilia” (Univ. Chicago Press, Chicago, 1956). 113. A. S. Romer and L. Price, “Review of the Pelycosauria,” Spec. Pap. Geol. Soc. Am., No. 28, 1–538 (1940). 114. B. S. Rubidge, “The Cranial Morphology and Paleoenvironment of Eodicynodon Barry (Therapsida, Dicynodontia),” Paleontol. Navors. Nas. Mus. Bleomfontein 4 (14), 327–402 (1984). 115. B. S. Rubidge, “A New Primitive Dinocephalian Mammal-like Reptile from the Permian of South Africa,” Paleontology 34 (3), 547–559 (1991). 116. B. S. Rubidge, “Australosyodon, the first primitive Anteosaurid Dinocephalian from the upper Permian of Gondwana,” Paleontology, 55, Part 3, 579–594 (1994). 117. B. S. Rubidge, Ed., “Biostratigraphy of the Beaufort Group (Karroo Supergroup),” S. Afr. Comm. Stratigr. Biostratigr. 1, 1–46 (1995). 118. N. Rybczynski, “Cranial Anatomy and Phylogenetic Position of Suminia getmanovi, a Basal Anomodont (Amniota: Therapsida) from the Late Permian of Eastern Europe,” Zool. J. Linnean Soc. 130, 320–373 (2000). 119. N. Rybczynski and R. R. Reisz, “Earliest Evidence for Efficient Oral Processing in a Terrestrial Herbivore,” Nature 411, 684–687 (2001). 120. R. R. Schoch, “A Complete Trematopsid Amphibian from the Middle Triassic of Germany,” J. Vertebr. Paleontol. 26 (1), 29–43 (2006). 121. H. G. Seeley, “Research on the Structure, Organisation and Classification of the Fossil Reptilia: 8. Further Evidence of the Skeleton in Deuterosaurus and Rhopalodon from the Permian of Russia,” Phil. Trans. R. Soc. London, Ser. B 185, 663–717 (1894). PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

CRANIAL MORPHOLOGY AND EVOLUTION OF PERMIAN DINOMORPHA (EOTHERAPSIDA) 122. A. G. Sennikov, “Early Thecodonts of Eastern Europe,” Tr. Paleontol. Inst. Ross. Akad. Nauk 263, 1–140 (1995). 123. A. G. Sennikov, “Evolution of the Permian and Triassic Tetrapod Communities of Eastern Europe,” Palaeogeogr., Palaeoclimatol., Palaeoecol. 120, 331–351 (1996). 124. C. A. Sidor and J. Welman, “A Second Specimen of Lemurosaurus pricei (Therapsida, Burnetiamorpha),” J. Vertebr. Paleontol. 23 (3), 631–642 (2003). 125. D. Sigogneau, Révision Systématique des Gorgonopsiens Sud-Africains (CNRS, Paris, 1970a). 126. D. Sigogneau, “Contribution a la conaissanse des Ictidorhinidés,” Palaeontol. Afr. 13, 25–38 (1970b). 127. D. Sigogneau and P. K. Tchudinov, “Reflections on Some Russian Eotheriodonts (Reptilia, Synapsida, Therapsida),” Palaeovertebrata 85 (3), 79–109 (1972). 128. D. Sigogneau-Russell, “Theriodontia I,” in Encyclopedia of Paleoherpetology, Ed. by P. Wellnhofer (Fischer, Stuttgart–New York, 1989), Part 17B, pp. 1–127. 129. C. Sullivan, R. R. Reisz, and R. M. H. Smith, “The Permian Mammal-like Herbivore Diictodon, the Oldest Known Example of Sexually Dimorphic,” Proc. R. Soc. London, Ser. B 270, 173–178 (2003). 130. M. V. Surkov, “The First Dicynodont from the Terminal Lower Triassic of European Russia, with Special Reference to the Evolution of the Masticatory Apparatus of These Therapsids,” Paleontol. Zh., No. 1, 76–82 (2005) [Paleontol. J. 39 (1), 72–78 (2005)]. 131. P. P. Sushkin, “On the Ethology of Dicynodon,” Dokl. Ross. Akad. Nauk 24, 11–12 (1922). 132. L. P. Tatarinov, “The Evolution of the Sound-Transmitting Apparatus in Lower Terrestrial Vertebrates and the Origin of Reptiles,” Zool. Zh. 37 (1), 57–74 (1958). 133. L. P. Tatarinov, “Order Therapsida: Therapsids,” in Fundamentals of Paleontology: Amphibians, Reptiles, and Birds, Ed. by A.K. Rozhdestvenskii and L.P. Tatarinov (Nauka, Moscow, 1964), p. 246 [in Russian]. 134. L. P. Tatarinov, “New Data on Ulemosaurus,” Paleontol. Zh., No. 2, 93–108 (1965). 135. L. P. Tatarinov, “New Data on the Formation in Theriodonts of the Additional Mammal-like Articulation,” Dokl. Akad. Nauk SSSR 166 (3), 749–752 (1966). 136. L. P. Tatarinov, “New Theriodonts from the Upper Permian of the USSR, in Upper Paleozoic and Mesozoic Amphibians and Reptiles of the USSR (Nauka, Moscow, 1968), pp. 32–46 [in Russian]. 137. L. P. Tatarinov, “Theriodonts of the USSR,” Tr. Paleontol. Inst. Akad. Nauk SSSR 143, 1–240 (1974). 138. L. P. Tatarinov, Morphological Evolution of Theriodonts and General Questions of the Phylogeny (Nauka, Moscow, 1976) [in Russian]. 139. L. P. Tatarinov, “A New Gorgonopian from the Upper Permian Deposits of the Vologda Region,” Paleontol. Zh., No 2, 97–104 (1977). 140. L. P. Tatarinov, “New Theriodonts (Reptilia) from the Late Permian Fauna of the Town of Kotelnich, Kirov Region,” Paleontol. Zh., No. 5, 76–82 (1999a) [Paleontol. J. 33 (5), 540–546 (1999)]. 141. L. P. Tatarinov, “A New Eotitanosuchian (Reptilia, Therapsida) from the Kazanian Stage (Upper Permian) of Udmurtia,” Paleontol. Zh., No. 6, 57–63 (1999b) [Paleontol. J. 33 (6), 636–642 (1999)]. 142. L. P. Tatarinov, “A New Gorgonopid (Reptilia, Theriodontia) from the Upper Permian of the Vologda Region,” Paleontol. Zh., No. 1, 70–78 (2000) [Paleontol. J. 34 (1), 68–76 (2000)]. PALEONTOLOGICAL JOURNAL

Vol. 42

No. 9

2008

995

143. P. K. Tchudinov, “Upper Permian Therapsids from the Ezhovo Locality,” Paleontol. Zh., No. 4, 81–94 (1960). 144. P. K. Tchudinov, “Family Eotitanosuchidae,” in Fundamentals of Paleontology: Amphibians, Reptiles, and Birds), Ed. by A.K. Rozhdestvenskii and L.P. Tatarinov (Nauka, Moscow, 1964a), pp. 247–249 [in Russian]. 145. P. K. Tchudinov, “New Data on Dinocephalians from the USSR,” Paleontol. Zh., No 2, 85–98 (1964b) [in Russian]. 146. P. K. Tchudinov, “New Dinocephalians from the Ocher Locality,” in Upper Paleozoic and Mesozoic Amphibians and Reptiles of the USSR (Nauka, Moscow, 1968), pp. 16–31 [in Russian]. 147. P. K. Tchudinov, “Early Therapsids,” Tr. Paleontol. Inst. Akad. Nauk SSSR 202, 1–230 (1983). 148. T. Thulborn and S. Turner, “The Last Dicynodont: An Australian Cretaceous Relict,” Proc. R. Soc. London, Biol. Sci. 270 (1518), 887–993 (2003). 149. H. von Trautschold, “Die Reste permischen Reptilien des Palaeontologischen Kabinets der Universität Kazan,” Nouv. Mem. Soc. Imper. Natur. Moscou 15, 1–39 (1884). 150. A. Tumarkin, “On the Evolution of the Auditory Conducting Apparatus: A New Theory Based on Functional Consideration,” Evolution 9, 221–243 (1955). 151. G. I. Tverdokhlebova and M. F. Ivakhnenko, “New Tetrapods from the Tatarian Stage of Eastern Europe,” Paleontol. Zh., No. 2, 122–126 (1994). 152. B. P. Vjuschkov, “On Gorgonopians from the Northern Dvina Fauna,” Dokl. Akad. Nauk SSSR 91 (2), 397–400 (1953). 153. B. P. Vjuschkov, “Suborder Theriodontia (Theriodonts): Superfamily Gorgonopsoidea,” in Fundamentals of Paleontology: Amphibians, Reptiles, and Birds, Ed. by A.K. Rozhdestvenskii and L.P. Tatarinov (Nauka, Moscow, 1964), pp. 258–285 [in Russian]. 154. G. N. Vysotskii, “Sketches on Soils and Ground Water Conditions,” Byull. Pochvoved., Nos. 1–2, 5–13; Nos. 3–4, 74–84; Nos. 5–8, 104–119 (1927). 155. D. M. S. Watson, “The Deinocephalia, an Order of Mammal-like Reptiles,” Proc. Zool. Soc. London 3 (2), 749–786 (1914). 156. D. M. S. Watson, “The Bases of Classification of the Theriodontia,” Proc. Zool. Soc. London 29 (1), 35–98 (1921). 157. D. M. S. Watson, “Dicynodon and Its Allies,” Proc. Zool. Soc. London 118 (3), 823–877 (1948). 158. D. M. S. Watson, “Palaeontology and Modern Biology,” (New Haven, 1951). 159. D. M. S. Watson, “The Evolution of the Mammalian Ear,” Evolution 7 (2), 159–177 (1953). 160. D. M. S. Watson, “On Bolosaurus and the Origin and Classification of Reptiles,” Bull. Mus. Comp. Zool. Harvard Coll. (Cambridge) 111 (9), 1–449 (1954). 161. D. M. S. Watson and A. S. Romer, “A Classification of Therapsid Reptiles,” Bull. Mus. Comp. Zool. Harvard Coll. (Cambridge) 114 (2), 37–89 (1956). 162. T. S. Westoll, “The Origin of the Tetrapods,” Biol. Rev. 18 (2), 78–98 (1943). 163. T. S. Westoll, “The Mammalian Middle Ear,” Nature 155, p. 114 (1945). 164. E. G. Wever and J. A. Vernon, “The Sensitivity of the Turtles Ear As Shown by Its Electrical Potentials,” Proc. Nat. Acad. Sci. USA 42 (5), 213–220 (1956).