Résumé :
L’étude des rongeurs Theridomorpha permet de suivre le déroulement d’une radiation adaptative pendant toute sa durée (Eocène moyen-Oligocène terminal), sur un territoire restreint à l’extrémité ouest de l’Europe Occidentale. Dans ce papier, la phylogénie de ce groupe est établie à partir d’une analyse cladistique reposant sur l’examen de 315 caractères (310 dentaires). Le groupe d’intérêt comprend 110 des 132 espèces (24 genres) de Theridomyoidea et deux genres encore inclus jusqu’ici dans les Reithroparamyinae qui rejoignent les Theridomorpha. Les groupes externes comprennent des Glires basaux, Cocomys, Tanquammys et 16 Ischyromyiformes. Un cadre phylogénétique robuste est produit, qui permet de clarifier la systématique des Theridomorpha. La position des Remyoidea (nov. sup.fam.) au sein des Ischyromyiformes, extérieure aux Theridomorpha, est confortée. Les Protadelomys et Tardenomys sont à la base des Theridomyoidea, avant la séparation en deux clades correspondant aux familles Pseudosciuridae et Theridomyidae. Les sous-familles sont consolidées : Pseudosciurinae et Sciuroidinae pour les Pseudosciuridae ; Issiodoromyinae, Oltinomyinae, Columbomyinae, Theridomyinae, auxquelles s’ajoute au moins une nouvelle sous-famille (Patriotheridomyinae), pour les Theridomyidae. La topologie des chrono-espèces (sensu Simpson), traitées antérieurement comme lignées évolutives, apparaît dans la plupart des cas sous forme de clades successifs dans lesquels les espèces sont le plus souvent arrangées de manière pectinée, émergeant dans l’ordre stratigraphique. L’analyse des caractères aux principaux nœuds permet de consolider les caractères diagnostiques des taxons et les tendances évolutives, ainsi que de discuter des divers parallélismes et convergences dans l’évolution des structures et patrons dentaires (e.g., émail des incisives unisérié chez les Issiodoromyinae et les Patriotheridomyinae, ou pseudo-multisérié chez les Blainvillimys les plus hypsodontes, les Protechimys et Archaeomys ; patrons dentaires téniodontes ; allongement des dents déciduales chez les Patriotheridomyinae, Issiodoromyinae et Theridomyidae ; sélénodontie ou lophodontie). Les dynamiques évolutives traduites par les changements morphologiques sont mises en relation avec les variations environnementales. Enfin, les implications biochronologiques de l’évolution des Theridomyoidea sont discutées.
Abstract:
The adaptive radiation of the rodents Theridomorpha occurred during a limited time window (middle Eocene to late Oligocene), on an area restricted to Western Europe. In this paper, the phylogeny of this group is established via a cladistic assessment of 315 morphological characters (310 dental). The group of interest encompasses 110 upon 132 species (24 genera) of Theridomyoidea, and two genera formerly included within the Reithroparamyinae, and which are included here within the Theridomorpha. The outgroups include basal Glires, Cocomys, Tanquammys and 16 Ischyromyiformes. A robust phylogenetic frame is produced, which allows clarifying the systematics of the Theridomorpha. Within the Ischyromyiformes, the Remyoidea (nov. supfam.) are set apart from the Theridomorpha. Protadelomys and Tardenomys represent the earliest offshoots of the Theridomyoidea, before the dichotomy between Pseudosciuridae and Theridomyidae. It supports the former subfamilies Pseudosciurinae and Sciuroidinae within the Pseudosciuridae; and for the Theridomyidae: the Issiodoromyinae, Oltinomyinae, Columbomyinae, Theridomyinae, with at least one new subfamily (Patriotheridomyinae). The topologies of the chronospecies (sensu Simpson), formerly considered as evolutionary lineages, appear in most cases as successive clades, in which the species are generally pectinately arranged and emerging in the stratigraphic order. The analysis of characters supporting the main nodes allow consolidating the diagnosic characters of the taxa and their evolutionary trends, as well as discussing the various cases of parallelism and convergence in the evolution of structures and dental patterns (e.g., uniserial incisor enamel for Issiodoromyinae and Patriotheridomyinae, or pseudo-multiserial for the most hypsodont Blainvillimys, Protechimys and Archaeomys; taeniodont dental patterns; lengthening of deciduous premolars for Patriotheridomyinae, Issiodoromyinae and Theridomyidae; selenodonty or lophodonty).
Evolutionary dynamics are analysed with respect to environmental changes. Finally, biochronological implications of the evolution of Theridomyoidea are discussed.
  

The family Didelphidae is represented by three species in the Sainte-Néboule site, phosphorites of Quercy (lower Oligocene, San Cugat's nivel): Amphiperatherium minutum (Aymard), Amphiperatherium sp. and Peratherium cuvieri (Fischer). Only the first  and  third species are abundant (88 and 97 pieces). This two populations are described. The marsupial fauna of the european lower Oligocene is not recognized in its entirety in this site. 

Based on isolated teeth two new scyliorhinid shark species, Scyliorhinus biformis nov. sp. and Scyliorhinus suelstorfensis nov. sp., are described from the Sülstorf Beds, early-middle Chattian, of Mecklenburg, northeastern Germany. They form part of a speciose assemblage of necto-benthic sharks and batoids which populated the warm-temperate to subtropical upper shelf sea of the south-eastern North Sea Basin. 

A classification should be both usable and useful,not too complex either in the amount of splitting or in the number of hierarchies involved, and not so simple as to give a false assurance of knowledge of relationships. Classifications are only possible because we do not have complete knowledge of the evolution of the organisms concerned because gaps in the record are necessary to allow the separation of the various taxa. Rodent classification is complicated by the large number of organisms involved and by the geat amount of parallelism that has taken place in the evolution of any and all features. If several independent features are characteristic of a certain taxon, should an effort be made to define the group on the basis of all the features, or should only one be selected as the determinant ? Unless the evolution of the several features was closely linked, the former solution will sooner or later lead to insurmountable problems. 

A new artiodactyl, Diacodexis antunesi n.sp., is described from the early Eocene of Silveirinha, Portugal. Comparisons are made with Diacodexis gazini GODINOT, 1978, D. varleti SUDRE et al., 1983, D. cf. varleti from Paris Basin sites, D. sp. from Dormaal and from localities in Spain and England, D. secans from North America and D. pakistanensis from Asia; affinities and evolutive tendencies are discussed. The presence of Diacodexis in the locality of Silveirinha confirms the very early Eocene age of the latter. As Diacodexis antunesi appears ta be more primitive than D. gazini from Rians (early Eocene of France), it lends corroboration to the interpretation (essentially based previously on condylarths) of the Silveirinha assemblage as the oldest Eocene fauna known in Europe and supports the hypothesis that early artiodactyls migrated from Europe to North America. 

The revision of the old collections of fossil birds from the “Phosphorites du Quercy” and the study of new material give the following results (Gruiformes, Cariamae) :  The humeri and most of the carpometacarpi described under the name Filholornis belong in Elaphrocnemus. The ulnae ascribed to Fïlholornis belong in Idiornis. Most of the post-cranial elements of the genera Elaphrocnemus and Idiornis are described and show great similarities with recent Cariamidae and Opisthocomidae, and fossil Bathornithinae.  A new genus and a new species, Oblitavis insolitus, are created in the sub-family Idiornithinae; two new species are described in the genera Elaphrocnemus (E. brodkorbz) and Idiornis (I. itardiensis), and the species Elaphrocnemus gracilis is transferred to the genus Idiornis.  The genus Propelargus Lydekker is transferred from the family Ciconiidae to Cariamidae.  A new generic name, Occitaniavis, is created for the species Geranopsis elatus, which belong in Cariamidae, while the type-species of the genus, Geranopsis hastingsiae, is a member of the Gruidae.  The affinities between the Quercy avifauna and the Neotropical one is emphasized by the occurrence of Phorusrhacidae, previously known only from the Cenozoic of South America and the Late Pliocene or Early Pleistocene of North America. Thanks to the material collected during the new excavations, the stratigraphical position of most of the species is stated precisely, and evolutionary lineages are outlined. This study shows that the suborder Cariamae, presently restricted to two South American genera, was already extremely diversified during the Eocene, and widespread in Europe and North America. 

Dichobunoid artiodactyls are described in this paper from the middle Eocene Geiseltal lignite deposits near Halle (Sachsen-Anhalt, Germany). The genera Eurodexis and Parahexacodus are established based on odontologica1 studies. The type-species are E. ceciliensis (FRANZEN & KRUMBIEGEL, 1981) and P. germanicus n. sp. from the "obere Mittelkohle" (see text), MP 13. The genera are referred to the new subfamily Eurodexeinae (Dichobunidae, Artiodactyla, Mammalia), which also contains the monospecific genus Eygalayodon SUDRE & MARANDAT, 1993. These new genera show many features similar to North American homacodontids and antiacodontids. They exemplify a high degree of diversity in European dichobunids. Moreover, the family Dichobunidae includes the Dichobuninae and the Hyperdichobuninae. The connection of the posthypocristid with the postentocristid, the shift of a medial to a lingual position of the hypoconulid on the lower molars and the development of a large, caniniform P1 are regarded as the principal tendencies of the eurodexeines. The entoconid is larger and more mesially situated compared to that of Diacodexis. The hypoconulid has a lingual position on a broad postcingulid. The elongation of the premolars and the presence of diastemata are considered associated with extension of the muzzle. This and the acute tubercles of the molars could indicate a limited degree of insectivory. Protodichobune cf. oweni and Diacodexis cf. varleli, first reported here from the "untere Unterkohle" (MP 11) of the Geiseltal, demonstrate the existence of forms in the basal part of the middle Eocene, that are direct descendants of lower Eocene ones. Eurodexis russelli n. sp. from MP 10 of Premontre is the most primitive representative of Eurodexis. A further part of this lineage could be Messelobunodon sp. from MP 11 of Messel (Germany). Another form from Premontre, referred to Eurodexeinae indet., represents the origin of the lineage to Parahexacodus germanicus n. sp. from the Geiseltal (MP 13). The identification of several lineages in MP 10 is evidence of an adaptive radiation during this time. Unfortunately the history of the earlier species of Diacodexis is not known well enough to exclude earlier diversifications. Our study shows the independent development of artiodactyls at the end of the lower Eocene in Europe, North America (and maybe in Asia) with much convergence. The question of the geographic origin of the genus Diacodexis is still open. 

Extensive sampling of the Siwalik deposits of the Potwar Plateau of northem Pakistan and from the Zinda Pir dome of central Pakistan has produced a fossil record of Miocene ctenodactylids that can be correlated with the paleomagnetic time scale. The early Miocene Prosayimys flynni (n. gen., n. sp.) is recognized as the first ctenodactylid in the Indian subcontinent. Prosayimys is ancestral to Sayimys. From the late early Miocene to the early late Miocene, there is an anagenetic succession of three species of Sayimys: S. cf. S. intermedius, S. sivalensis, and S. chinjiensis (n. sp.). Sayimys chinjiensis gave rise to the late late Miocene S. perplexus. A second lineage is represented by Sayimys minor, S. sp. A, and S. sp. B. 

Based on material found in about 15 localities the relationships of the genera Microparamys and Glirarus have been studied. One new genus, two subgenera and three species [Microparamys (Sparnacomys) chandoni n. subgen. and n. sp., Microparamys (Pantrogna) russelli n. subgen., Eoglirarus wildi n. gen. and n. sp., Gliravus meridionalis n. sp.] as well as the publication
of numerous new facts concerning species previously reported, support the phyletic scheme proposed. The latter shows that the origin of the Gliravinae is to be sought among the very small and still rather poorly known Microparamys species of the early Eocene. Gliravus: robiacensis can be considered as the common ancestor to different lineages not only of Glirarus but also of modern genera (Peridyromys, Glirudinus and Microdyromyx), at the origin of which Peridyromys micio, although difficult  to interpret, occupies a similar place.

The stratigraphic conclusions permit more detail in the chronologie succession of the localities studied. The paleoecologic and biogeographic aspects lead one to the problem of the oligocene "Grande Coupure" through the study of this group.

  

Two somewhat “odd” taxa of theridomyid rodents, one formerly known (Bernardia marandati Vianey-Liaud, 1991) and the other new (Idicia vidalenci gen. et sp. nov.) are discussed from a taxonomical and taphonomical perspectives. These two rodents were found at Lavergne, a late middle Eocene (MP16) “phosphatière” from the Quercy (Southwest France). The genus Bernardia, being preoccupied by a scale insect (Bernardia Ashmead, 1881), is here renamed Burgia. We benefit from this nomenclatural change to describe additional new dental specimens of this patriotheridomyine species, including a previously undescribed locus (P4). The other theridomyid from Lavergne, Idicia vidalenci gen. et sp. nov., so far documented by a mandible preserving two teeth (m2-m3) is a new taxon of peculiar occlusal morphology, and whose subfamilial affinities remain unknown. These two peculiar theridomyids recorded at Lavergne are found nowhere else, whether in coeval localities in Quercy or elsewhere in Western Europe. We discuss the possible causes of their unique presence at Lavergne.
  

The petrosal of the Paulchoffatiidae HAHN, 1969 is described and compared with that of younger multituberculates and of other Mesozoic mammals. The "Morrison petrosal", described by Prothero (1983), is also discussed; it probably belongs to the multituberculates. The reconstruction of the ventral side of the Paulchoffatiinae-skull, given by Hahn in 1987, is completed by addition of the otic and the occipital region. 

Extensive collecting in previously known and new Middle Eocene rodent localities in the Subathu Group of the Rajauri and Reasi districts, Jammu and Kashmir (northwest Himalaya, India) has yielded over 500 isolated cheek teeth, numerous incisors and cheek tooth fragments, and two mandibular fragments with M/1-M/3. An analysis of this additional material and a restudy of that reported earlier from the Metka (Rajauri) area has revealed the presence of diverse Middle Eocene ctenodactyloid rodents in India. The assemblage comprises six genera and twelve species, viz.., Bírbalomys woodi, B. ibrahimshahi, B. sondaari, Basalomys vandermeuleni, B. ijlsti, B. lavocati, Chapattimys wilsoni, C. debruijni, Gumbatomys asifi, cf. Advenimus bohlini, cf. Petrokoslovia sp. indet. 1 and cf. Petrokoslovia sp. indet. 2. Most species of this assemblage are endemic to the Indian subcontinent and are referred to the Family Chapattimyidae. However, at least three taxa, viz.., cf. Advenimus bohlini, cf. Petrokoslovia sp.indet. 1 and cf. Pezrokoslovia sp. indet. 2 show close affinities with their Central and Eastern Asiatic contemporaries and are attributed to the Family Yuomyidae. The dentitions of most of the Middle Eocene rodents from India are described here in better detail thus providing more precise characterization than was available earlier. The taxonomie status of Basalomys, formerly a subgenus of Birbalomys has been raised to that of a genus. Basalomys vandermeuleni is here designated as the type species of Basalomys in place of B. ijlsti. The rodent assemblage from the Subathu Group is specifically similar to that from the coeval Kuldana Formation of Pakistan. The absence of Birbalomys ibrahimshahi and cf. Advenimus bohlini in Pakistan is related to sampling bias. The identifications of the Early Eocene rodents from Barbara Banda, Pakistan have been reexamined and the familial assignments of some of the Early and Middle Eocene rodents from Central and Eastern Asia have been discussed and reviewed in the light of recent discovery of early rodents in Mongolia.

  

The Oligocene deposits of the Ulantatal area in Inner Mongolia (China) contain among the richest mammalian faunas from Asia. To date, only some parts of the rodent faunas have been described. Here, we propose to review the rodent faunal lists for each site, including the description of a few new rodent specimens. We describe three additional rodent species: the Cylindrodontidae Anomoemys lohiculus, the Eomyidae Asianeomys sp., and the Dipodidae Litodonomys huangheensis. This study allows us to constrain the stratigraphic range of Anomoemys lohiculus, which ranged from the late Early Oligocene to the early Late Oligocene in this area. Asianeomys sp. and Litodonomys huangheensis are dated from the latest Oligocene. These Oligocene deposits consist now of more than 70 species of mammals if we include the fauna from Kekeamu. This latter corresponds to the basal part of the Ulantatal Formation and could be dated biochronologically from the earliest Oligocene. When compared to the faunas from the Valley of Lakes in Central Mongolia, the Ulantatal faunas present a great majority of rodents, and this difference can be partly explained by sampling and description biases regarding macro-mammals. This study also shows that variations existed between Inner and Central Mongolia, especially regarding the composition of the rodent paleocommunities. However, the assessment of their evolutionary history in this part of Asia with respect to the important climate and environment changes, require further precisions and more material than current data allow. 

The Las Águilas site near Porvenir de Jalpa, Coahuila, Mexico, is extremely rich in tetrapod remains comprising both bones and trackways of several dinosaur taxa of late Campanian age. Within a 50 m thick section we identified at least nine layers with dinosaur bone assemblages. In one of these the dinosaur bones are associated with remnants of eusuchian crocodilians, turtles, plesiosaurs, pterosaurs, tyrannosaurids, dromaeosaurids, parksosaurid, hadrosaurids, ceratopsids, and ankylosaurs. This layer is also rich in coprolites of turtles, crocodilians and likely theropods, thus providing evidence for the wealth of Late Cretaceous vertebrate life in the area. 

The description of new material of the Upper Eoœne of France, Southem Germany and Spain together with the restudy of the material described by Thaler (1966b) as Pairomys allows to confirm the validity of the genus Ectropomys up to know controversally discussed. As a consequence the subfamily Remyinae and Oltinomyinae can be defined precisely. The study of the dental characters shows that the Remyinae in fact belong to the Theridomyidae and represent a derived evolutionary unit within this family. The Oltinomyinae together with the Theridomyinae and Issiodoromyinae very probably form a monophyletic unit. 

The locality of Montredon has provided 8 species of camivores, often little documented, that are discribed and depicted. All the remains are fragmentary and generally badly preserved which lowers very much the possible precision of the taxonomic study; this one has however allowed the creation of a new subspecies. The most richlypdocumented forms are an ursid (Indarctos) and a felid (Machairodus). Issuing often from a westem Europe evolution, the carnivores are well inserted between those of the better known faunas of the "Classical Pontian" and of the lower Vallesian and this situation fits also well with their stratigraphic level, that cannot however be more accurately defined with them alone.
  

Previous chronological arrangements of South American Quaternary land mammal faunas are appraised on the basis of current geological and paleontological data. Three South American late Pliocene-Pleistocene land mammal ages are conventionally recognized, from oldest to youngest, the Uquian, Ensenadan, and Lujanian ; all are defined on Argentine faunas.

     The Uquian is based fundamentally and historically on the fauna from the Uquía Formation in Jujuy Province, northwestern Argentina. Important known formations in Argentina yielding Uquian Age faunas include the sub-surface Puelche Formation (or Puelchense) near the city of Buenos Aires, and the Barranca de Los Lobos and Vorohué Formations between Mar del Plata and Miramar, Buenos Aires Province. A tentative subdivision is propos-ed for the Uquian into three subages based on knowledge of the Mar del Plata-Miramar sequence, from oldest to youngest, the Barrancalobian, Vorohuean, and Sanandresian. In Argentina the Uquian is presently marked by the first known record of Scelidodon, Hydrochoeropsis, Ctenomys, Canidae, Ursidae, Gomphotheriidae, Equidae, Tapiridae, Camelidae, Cervidae, and the last known record of Thylatheridium, Thylophorops, Dankomys, Eumysops, Pithanotomys, Eucoelophorus, Hegetotheriidae, Sparassocynidae, and Microtragulidae.

The Ensenadan Age is based on the fauna from the Ensenada Formation near the city of Ensenada, Buenos Aires Province. In Argentina the Ensenadan is marked by the first known record of Lomaphorus, Neothoracophorus, Plaxhaplous, Cavia, Lyncodon, Lutra, Galera, Smilodon, Dicotyles, Lama, Vicugna, the last known record of Orthomyctera, and the only known record of Brachynasua.

     Typícal beds of late Lujanian Age in Argentina consist of fluvial deposits occupying stream channels, and shallow basins, often incised into beds of early Lujanian (i.e. Bonaerian of early workers) and Ensenadan Age. The Lujanian Age is based on a fauna from beds along the Rio Luján, about 65 km west of the city of Buenos Aires, Buenos Aires Province. The Lujanian in Argentina is marked by the first record of Equus, Chlamyphorus, and Holochilus, and the last record of Megatherioidea, Glyptodontoidea, Arctodus (=Arctotherium), Smilodon, Litopterna, Notoungulata, Proboscidea, Equidae, Morenelaphus, and Palaeolama.

   These land mammal ages are often difficult to recognize in other South American countries. The compositions of South American Pleistocene faunas vary with the environment. Some taxa were widely distributed in fossil deposits throughout the continent, but their occurrences need not reflect synchroneity. This is a result of changing climates and habitats in time. Consequently, proposed intracontinental correlations need confirmation based on magnetostratigraphy and a radioisotope time scale. Paleontologic characterizations of these land mammal ages (i.e. first and last record, and guide fossils) are useful for much of Argentina, but extensions to most of the other parts of South America are at best tenuous.

The majority of known non-Argentine Pleistocene faunas are believed to be Lujanian in age. Possible non Argentine early Pleistocene (Uquian) faunas include Ayo Ayo and Anzaldo in Bolivia, and Cocha Verde in southern Columbia. A possible middle Pleistocene (Ensenadan or early Lujanian) fauna is the Chichense of Ecuador. Paleomagnetic and radioisotopic date (MacFadden et al., 1983) clearly indicate that the greater part of the Tarija fauna (Bolivia) is Ensenadan in age.

  The end of the Pleistocene and beginning of the Holocene in South America is marked by extinction of nearly all large mammalian herbivores and their specialized large predators. Radiocarbon age determinations suggest that large scale extinctions of megafauna occurred between 15,000 and 8,000 yrs. B.P. (years before present). 

The exploitation of new early Eocene localities in the Paris Basin has resulted in the collecting of  numerous mammalian remains, among which are about 300 isolated teeth representing the rodents. They belong, for the most part, to the paramyid group. Only the latest level of the early Eocene has yielded rodents belonging to the pseudosciurid group. The paramyids, the object of this study, are represented by at least 5 genera and 10 species; they are distributed among 4 clearly dilferentiated subfamilies : Paramyinae Simpson 1945, Pseudoparamyinae Michaux 1964, Ailuraviínae n. subf., Microparamyinae Wood1962.
It results from this study that the principal types of rodents in the middle and late Eocene of Europe are clearly related to the forms described here: Plesiarctomys Bravard 1850 is related to Pseudoparamys Michaux 1964, Ailuravus Rütimeyer1891 to Meldimys n. gen., Gliravus hammeli Thaler 1966 to Microparamys nanus (Theilard1927), and some Masillamys Tobien 1954 to Microparamys russelli Michaux 1964 and to M. sp. 1. Gliravus and Masillamys are the oldest representatives of the Gliridae and the Pseudosciurídae, respectively, families which will progressively replace the Paramyidae in the course of the middle and late Eocene.
Two stages can be recognized in the history of te European Paramyidae fauna during the course of the early Eocene: the older is characterized by the Mutigny fauna, the younger by the Cuis fauna.
The Mutigny fauna appears already rather diversified and does not completely correspond, point by point, to the homologous fauna of North America. A geographic differentiation seems to have been manifested rapidly, unless the fauna established in Europe was already a little different from that which established itself in North America. 

A cheek tooth recently unearthed in the Lopar Sandstone unit, of late Eocene age, in the northern part of Rab Island, Croatia, is one of the very few Eocene mammalian remains found in the Adriatic area. Thorough comparison of this tooth with those of Old-World Palaeogene mammalian orders suggests that it is a M3 belonging to an embrithopod afrothere. The specimen is referred to as Palaeoamasia sp. This genus was formerly known only in Eocene deposits of Anatolia but with close relatives in Romania among Palaeoamasiinae. The geographical distribution of this subfamily perfectly matches the recently-named Balkanatolian landmass, which experienced in-situ evolution of endemic mammals prior to the Grande Coupure event that occurred around the Eocene–Oligocene transition. This last event is characterised by massive Asian immigration in Western Europe and the supposed extinction of many endemic Central and Western European mammals, including Palaeoamasiinae.
  

There is only one suid known in the Upper Miocene of Montredon (Hérault): Microstonyx (Limnostonyx nov. subgen.) antiquus (KAUP). It is differenciated from Microstonyx major by the presence of upper and lower canines which are considerably longer and biger. Its presence at Montredon corroborates the palustrine habitat for the species.