Mesozoic Fishes 2 – Systematics and Fossil Record

Barbara J. STAHL: Mesozoic holocephalians
[pp. 9-19, 4 figs.]

Almost all the Mesozoic holocephalians can be characterized either as myriacanthoids or chimaeroids. The interrelationship of the Mesozoic holocephalians and their connections to the Paleozoic holostylic chondrichthyans is problematic. Evidence from dentitions and tooth structure is difficult to interpret; body form and skeletal design is superficially similar but poorly known. Reduction in the number of tooth plates comprising the dentition can be predicated on the basis of a morphologic series and the apparent shift in the direction of tooth plate growth can be explained, but a phylogenetic progression from a tooth plate with a base of laminar tissue and a crown of orthotrabeculine to one of chimaeroid design, lacking both these features, is purely hypothetical. The skeletal evidence, especially cranial structure, indicates a difference in adaptive solutions to functional requirements. The myriacanthoid most similar to the chimaeroids in cranial structure is Chimaeropsis; the skull in that form cannot be studied further as the unique specimen has been lost. Myriacanthoid remains are still too sparse to permit speculation on relationships within this group of holocephalian fishes. More can be said about the evolution of the chimaeroids, although here, too, important parts of the record are missing. The availability of extant representatives makes it possible to affirm with some confidence the progressive reduction of tritoral tissue in the tooth plates and the increasingly sectorial nature of the dentition in these holocephalians. The classification of the known Mesozoic chimaeroids must be rethought: newly discovered specimens suggest that several of the generic assignments made by 19th century investigators are untenable. Genera were often based on tooth plates alone and differences in appearance due to age and wear may have led to the erection of spurious taxa. Discovery of associated dentitions will allow prunning of the nomenclature.

Alberto Luis CIONE: First report of a Jurassic ray outside of Europe
[pp. 21-28, 4 figs.]

A "rhinobatid" ray was found in Tithonian beds of the Vaca Muerta Formation at Cerro Yrigoyen, near Las Lajas, Provincia del Neuquén, southwestern Argentina.

The fossil consists of a partial skeleton including a great part of the skull, branchial arches, left pectoral fin, and pectoral girdle. The cartilages are mostly preserved in three dimensions.

The specimen is relatively small (total length less than 0.5 m). The pectoral fins are moderately expanded. The synarcual is large and incorporates several vertebrae. The ventral placoid scales are small, flat, and polygonal.

This is the first Jurassic ray reported outside Europe and the first Jurassic selachian in South America. Previously, Jurassic rays were known from the Tethyan sea (Germany, England, Spain, and France). The present record reveals that rays were also distributed in the southern hemisphere during the Late Jurassic.

Armin LEIDNER & Detlev THIES: Placoid scales and oral teeth of Late Jurassic elasmobranchs from Europe
[pp. 29-40, 3 figs.]

Thirty-nine specimens of articulated skeletons of sharks and rays from the Lithographic Limestones of Southern Germany and France (Upper Jurassic, Kimmeridgian and Tithonian) were sampled in order to document placoid scale and dental morphologies of Late Jurassic elasmobranchs. The study reveals that the systematics of Late Jurassic elasmobranchs is in urgent need of revision. The material contains two new, as yet undescribed species: Synechodus sp. nov. and Squatinidae gen. et spec. nov. Squatina speciosa (MEYER 1856) is a junior synonym of Squatina alifera (MÜNSTER 1842). The previously unknown or only little known dental morphologies of Squatina alifera (MÜNSTER 1842), Phorcynis catulina THIOLLIÈRE 1854, Palaeoscyllium formosum WAGNER 1857, and Asterodermus sp. are illustrated. The interrelationships of Asterodermus platypterus AGASSIZ 1843, Belemnobatis sismondae THIOLLIÈRE 1854, and Spathobatis bugesiacusTHIOLLIÈRE 1854 remain obscure.

William E. BEMIS & Lance GRANDE: Development of the median fins of the North American paddlefish (Polyodon spathula), and a reevaluation of the lateral fin-fold hypothesis
[pp. 41-68, 7 figs., 1 tab.]

To understand more about the structure of median fins of "lower actinopterygians" (that is, actinopterygians outside of Neopterygii), we investigated development of median fins in the North American paddlefish, Polyodon spathula (Acipenseriformes: Polyodontidae). We observed that dorsal and ventral median fin-folds form during the embryonic period, and are well-developed by hatching (stage 36). Fin-webs of both the dorsal median fin-fold and ventral median fin-fold greatly narrow at the posterior tip of the body, so that the notochordal axis reaches the most caudal point of the body. The three median fins (dorsal, anal and caudal) differentiate rapidly from the median fin-folds during the yolk-sac larval period (stages 36 to 46). The order of differentiation is dorsal, anal and caudal, which reflects the general gradient of development from anterior to posterior. Myotomes extend first into middle portions of dorsal and anal fins. Cartilage condensations that will form radial elements and bundles of actinotrichia that will contribute to fin-rays also develop first in the middle portions of these fins. Radial skeletal elements of dorsal and anal fins differentiate in this order: middle radials, proximal radials, and finally distal radials. Juveniles have only slightly fewer radials in the dorsal and anal fins than do adults. Thus, fin endoskeletons are essentially fixed by metamorphosis, that is, about 39 days old, when fin shapes approximate those of adults, and adult coloration and feeding mode are achieved.

Pectoral and pelvic fins of Polyodon share a skeletal design (based on radial elements) with dorsal and anal fins. In contrast, the endoskeleton and exoskeleton of the caudal fin differ from those of the dorsal, anal and paired fins. Unlike other fins, myotomal extensions do not contribute to either the epichordal or hypochordal lobes of the caudal fin. Dorsally, the epichordal lobe develops from a portion of the dorsal median fin-fold, and its endoskeleton consists of a series of supraneural elements and neural arches; in contrast, such elements do not participate in formation or support of the dorsal fin. Fulcra form along the leading edge of both the epichordal and hypochordal lobes of the caudal fin; in contrast, fulcra are absent from dorsal, anal and paired fins. Ventrally, the hypochordal lobe forms from a portion of the ventral median fin-fold, and its endoskeleton consists chiefly of a series of hypural bones that is serially homologous with hemal arches and spines; in contrast, such elements do not participate in formation or support of the anal fin. The caudal fin also has a few so-called distal caudal radials of uncertain homology.

A small preanal portion of the ventral median fin-fold persists between the anus and yolk-sac throughout the yolk-sac larval period. Neither myotomes nor cartilage condensations contribute to this preanal fin, and it is lost during further development. Pelvic fin-buds develop later (stage 38) than do median fins (<stage 36). Median fin-folds are never continuous with the pelvic fin buds. A wide stretch of flank separates the pelvic and pectoral fin-buds, with no trace of a continuous fin-fold between them.

For more than a century, the structure and development of fins of Acipenseriformes have been studied by authors primarily interested in evolution of gnathostome fins. Fins of Acipenseriformes and other gnathostomes have been interpreted through many hypothetical (that is, non-empirical) models, which now are entrenched in the literature. We reevaluate some of the conclusions and extrapolations of these models, and concepts related to the lateral fin-fold hypothesis. We find little developmental or phylogenetic support for most of these ideas, and urge a retreat from such idealistic morphology.

Eric J. HILTON & William E. BEMIS: Skeletal variation in shortnose sturgeon (Acipenser brevirostrum) from the Connecticut River: Implications for comparative osteological studies of fossil and living fishes
[pp. 69-94, 13 figs., 4 tabs.]

To better understand an extreme case of individual skeletal variation, we examined in detail a new collection of thirteen shortnose sturgeon (Acipenser brevirostrum) from a single, well-characterized population in the Connecticut River of western Massachusetts. We discovered and analyzed four types of variation: 1) lateral line and anamestic ossifications of the dermal skull roof; 2) development of scutes; 3) chondral ossifications in the scapulocoracoid; and 4) chondral ossifications in the neurocranium. Variation within the skull roof is extreme, even more than that of other non-teleostean actinopterygians studied to date. Each individual in our sample displays a unique pattern of ossification in the skull roof. Scute formation is variable, with some individuals possessing poorly developed or incomplete scutes. However, the number of individuals with such extreme variation in scute morphology was small relative to the number showing other forms of skeletal variation. The degree of ossification of the scapulocoracoid differs, even between animals of similar fork-length and estimated ages. Some right-left differences in scapulocoracoid ossifications also occur. The number of discrete ossification centers within the neurocranium also varies greatly and is not correlated with size or age. This study offers several insights for studies of fossil and living fishes. First, it illustrates an extreme example of individual skeletal variation. Second, it shows how little is known about the regulatory mechanisms that control patterns of chondral and dermal ossification during the ontogeny of an individual. Third, it indicates potential problem areas in determining the homologies of skeletal elements. These insights suggest caution in the interpretation of small samples in comparative and descriptive osteological studies, both of fossil and living fishes.

Christopher FIELITZ, J. D. STEWART & Joan WIFFEN: †Aethocephalichthys hyainarhinos gen. et sp. nov., a new and enigmatic Late Cretaceous actinopterygian from North America and New Zealand
[pp. 95-106, 7 figs.]

A new actinopterygian, Aethocephalichthys hyainarhinos gen. et sp. nov., is described. It is Late Cretaceous in age and found in the Niobrara, Eutaw, and Pierre Shale formations of North America, and in the Maungataniwha Sandstone Formation of New Zealand. Almost all specimens consist solely of the neurocranium. It is rectangular in shape with a blunt snout and characterized by the fusion of all the bones of the neurocranium. The bones of the roof and snout are covered with a dermal pattern of polygons, many of which have a central pore. Each pectoral fin-ray has a shallow channel with pits along the channel. This species shares a number of primitive characters with various actinopterygian groups such as an ossified interorbital septum. Potential characters that could place Aethocepalichthys within the teleosts are inconclusive due to the bone fusion. These include a deep posterior myodome and an unpaired vomer. We conclude that this fish is an incertae sedis actinopterygian.

Sylvie WENZ: †Pliodetes nigeriensis, gen. nov. et. sp. nov., a new semionotid fish from the Lower Cretaceous of Gadoufaoua (Niger Republic): phylogenetic comments
[pp. 107-120, 10 figs.]

†Pliodetes nigeriensis, a new genus and species of semionotid fish is described from the Lower Cretaceous (Aptian) of Gadoufaoua, Niger Republic. It presents a unique combination of primitive and derived characters. For instance: 1) posterior extremity of premaxillary ascendent process superficial, occasionally ornamented, and taking part of the dermal cover of the skull roof; (2) reduction of the dermal ossification of the lateral face of the mandible; and (3) presence of ossified vertebral centra. The morphology and number of ossifications of the lateral face of the mandible and the type of dentition allow discrimination of a group formed by †Semionotus, †Paralepidotus, and †Lepidotes within the semionotids. This group is characterized by the presence of a coronoid process and of three ossifications on the lateral face of the mandible (dentalosplenial with a long posterior process, angular, and surangular). A second group corresponding to †Pliodetes gen. nov. and †Araripelepidotes is characterized by the absence of coronoid process and by the reduction in the number of ossifications on the lateral face of the mandible (two in †Pliodetes, dentalosplenial and angular; only one, a boomerang-shaped dentalosplenial in †Araripelepidotes).

Gloria ARRATIA & Andrea TINTORI: The caudal skeleton of the Triassic actinopterygian †Prohalecites and its phylogenetic position
[pp. 121-142, 13 figs., 1 tab., 1 app.]

The caudal fin of †Prohalecites is hemiheterocercal, both with an epaxial and a hypaxial basal fulcra, and fringing fulcra, and with fewer than 19 principal caudal fin-rays. The caudal skeleton has irregularly shaped and expanded preural and ural neural arches that usually keep their neural spines. Individuals with complete neural arches (including their spines) do not have epurals. In contrast, individuals with incomplete preural and ural arches bear three to five epurals. Similar preural and ural neural arches are present in †caturids and in †semionotiforms, e.g. †Dapedium and †Tetragonolepis. The expanded and elongated ural neural arches produce a bony block increasing the stiffness of the tail during locomotion (a function similar to that of the calcified ural neural cartilages in †caturids and of the uroneurals in teleosts). According to the morphology of these elements and to the phylogenetic hypothesis, the last expanded preural and ural neural arches of †Prohalecites and †semionotiforms and the modified preural and ural neural arches (or uroneural-like elements) of †pachycormiforms are not uroneurals (paired elongated ural neural arches) like those of Teleostei. However, all of them result from modifications of the last caudal neural arches.

†Prohalecites possesses an interesting mosaic of primitive and advanced characters, some of which have been previously interpreted as synapomorphies of Teleostei, e.g., presence of a posteriorly expanded frontal bone of distinct shape, a mobile premaxilla, a propterygium fused to the first pectoral ray, and fewer than seven epurals. It shares with the †Macrosemiidae the presence of a ventral preopercular bone, which has been previously interpreted as a unique character of the family. Among the generalized characters, †Prohalecites has two vomerine tooth plates, a suprangular bone, one hypohyal, a polyural caudal skeleton, and fringing fulcra on all fins. Autapomorphies of †Prohalecites are the presence of large nasal bones sutured to each other and the loss of body scales. †Prohalecites shows different phylogenetic positions among non-teleostean fishes when the outgroups are changed, but it is still more advanced than Amiidae and more primitive than †aspidorhynchiforms, †pachycormiforms, and †pycnodontiforms.

Detlev THIES & Annette HERZOG: New information on †Dapedium LEACH 1822 (Actinopterygii, †Semionotiformes)
[pp. 143-152, 6 figs.]

One specimen each of Dapedium caelatum QUENSTEDT 1858, D. pholidotum (AGASSIZ 1832), and D. punctatum AGASSIZ 1835 from the Posidonia shale (Lower Jurassic, lower Toarcian) of Southern Germany were studied. The study revealed the existence of an independent intercalar (D. punctatum), a paired vomer (D. caelatum), an independent dentary and splenial (D. caelatum, D. pholidotum), and branchial tooth plates (D. caelatum, D. pholidotum, D. punctatum) in the skeletons of late Liassic specimens of the genus Dapedium. Also, in the specimen of D. pholidotum, the coronoids constitute part of the formation of the lateral cover of the lower jaw. In this specimen the marginal row of teeth of the mandible is formed mostly by the lateral coronoid teeth.

J. R. NURSALL: The family †Mesturidae and the skull of pycnodont fishes
[pp. 153-188, 23 figs., 2 tabs.]

The family †Mesturidae comprises four genera: Mesturus WAGNER 1862; Micropycnodon HIBBARD & GRAFFHAM 1945; Arduafrons FRICKHINGER 1991, an incompletely described genus, which has been mentioned in the literature under different names; and Paramesturus TAVERNE 1981. Descriptions of the following species are emended: Mesturus verrucosus WAGNER; Mesturus leedsi WOODWARD; and Micropycnodon kansasensis(HIBBARD & GRAFFHAM). The new species Micropycnodon gaynaisensis is described. A more complete description is provided for Arduafrons prominoris FRICKHINGER. Paramesturus is commented upon briefly. Micropycnodon gaynaisensis has a less distorted skull than M. kansasensis, which allows a more accurate description of the dermal skull of this genus. Dermatocrania are compared among †Mesturidae, then within †Pycnodontiformes as a whole. The dermal skull of gyrodontoid fishes has more units than that of pycnodontoid fishes. The extra units include median nuchals, interfrontals and dermethmoids as well as paired marginals and lateral dermethmoids, all of which have been given different names by different authors. Homologization of the extra bones with dermal skull bones in other actinopterygians is difficult. Standardization of terminology for dermatocranial bones in pycnodonts is attempted. A nearly complete neurocranium of Mesturus sp. is described. Bones described include frontal, parietal, dermopterotic, dermosupraoccipital, orbitopterosphenoid, basisphenoid, prootic, sphenotic, parasphenoid, pterotic, opisthotic, intercalar, basioccipital, exoccipital, epioccipital, supraoccipital, and the foramina, grooves, chambers and facets that distinguish them. Neurocrania of other pycnodonts are compared. Synapomorphous dermal structures of pycnodont fishes include the dermosupraoccipital, nuchals, and marginals. The supraoccipital of pycnodonts is homologous to that of teleosts (including pholidophorids). It is concluded that the arrangement of elements in the neurocranium of pycnodonts is within the range possible for teleost fishes.

J. Ralph NURSALL: The pycnodontiform bauplan: The morphology of a successful taxon
[pp. 189-214, 16 figs., 2 tabs.]

The Pycnodontiformes were a chiefly Mesozoic, sister-group to Teleostei. They were mostly deep-bodied, laterally compressed fishes, distinguished by a suite of morphological characters, from primitive to highly derived states, relative to other halecostomes. The characters include features of teeth, upper and lower jaws, opercular series, pterygoquadrate arcade, parasphenoid, neurocranium, sensory canals, vertebral column, caudal skeleton, and ribs, as well as a number of derived characters shared by Pycnodontidae, such as reduction in scales, parietal peniculus, temporal fenestra, elaborate zygapophyses, pharyngeal teeth, etc. A composite pycnodont has been drawn to illustrate the features of the order and a tabulation made of important apomorphies. It is concluded that many of the characters of pycnodont fishes are neotenic, which causes problems in establishing the phylogenetic position of the order. It is suggested that the possibility exists to replicate character states similar to those in pycnodonts by the manipulation of homeotic genes in fish currently used in ontogenetic studies.

Jürgen KRIWET: Pycnodont fishes (Neopterygii, †Pycnodontiformes) from the Lower Cretaceous of Uña (E-Spain) with comments on branchial teeth in pycnodontid fishes
[pp. 215-238, 12 figs., 1 tab.]

The Lower Cretaceous (upper Barremian) locality of Uña in the Province of Cuenca (E-Spain) yielded a diversified vertebrate fauna including three pycnodont fishes. For the first time Anomoeodus is reported with a new species, A. nursalli sp. nov., from the Lower Cretaceous of Spain. The material consists of isolated and disarticulated specimens. It is possible to prove the presence of branchial teeth in Anomoeodus and to show the variability in tooth morphology and arrangement of the new species. Branchial teeth are also reported for the first time in the Brasilian pycnodontid Iemanja WENZ 1989. The morphology, location, and wear patterns of branchial teeth indicate an important role in feeding behavior. Two other pycnodonts are represented by vomerine dentitions only. One species is referred to Coelodus. The second one represents a species of uncertain affinities. This species shares dental characters with Coelodus and Eomesodon.

Paulo M. BRITO: Description of Aspidorhynchus arawaki from the Late Jurassic of Cuba (Actinopterygii: Halecostomi)
[pp. 239-248, 5 figs.]

Aspidorhynchus arawaki from the Jagua Formation, Pinar del Río Province (Oxfordian of Cuba), is described on the basis of acid prepared material. The Cuban species is distinguished from other species of the genus by the length of the premaxilla, the size of the maxillary teeth, the dimensions of the predentary, and the presence of a large median tooth in the posterior part of the predentary.

Paulo M. BRITO: The caudal skeleton of aspidorhynchids (Actinopterygii, Halecostomi): phylogenetic implications
[pp. 249-264, 7 figs.]

Redescription of the caudal skeleton of some aspidorhynchids suggests that the family Aspidorhynchidae belongs within the Teleostei. This assigment is based on the presence of ural neural arches modified as uroneurals and the reduction of number of hypurals.

Within aspidorhynchids, Aspidorhynchus and Belonostomus share one synapomorphy (absence of neural arch and spine in the first ural centrum). Vinctifer has the first two ural centra fused into a single element.

The short and squat "uroneurals" found in Belonostomus have a different origin than that of the uroneurals: the first "uroneural" originates from modification of preural neural arches; therefore, it is not an uroneural. The following elements are modified ural neural arches and they are uroneurals. Consequently with such a conclusion, the aspidorhynchids are considered teleosts.

Gloria ARRATIA: The monophyly of Teleostei and stem-group teleosts. Consensus and disagreements
[pp. 265-334, 28 figs., 2 tabs.]

The relationships among Jurassic teleosts from Europe, and Middle and South America, the ?Early Cretaceous osteoglossomorph †Lycoptera, the Early Cretaceous ostariophysan †Gordichthys and clupeomorph †Santanaclupea, the Cretaceous and Tertiary clupeomorph †Diplomystus, the Cretaceous "salmoniforms" †Erichalcis and †Humbertia, and the extant Elops, Megalops, Hiodon, Heterotis, Chanos, Opsariichthys, Denticeps, Engraulis, Esox, Umbra, Oncorhynchus, and Thymallus are evaluated based on 196 unweighted morphological characters by using cladistic principles. †Aspidorhynchiforms and †pachycormiforms were included as part of the ingroup to test whether these fishes are teleosts or not.

The new phylogenetic hypothesis, based on both fossil and extant teleosts, shows that †Pholidophorus bechei is more primitive than †Leptolepis coryphaenoides, which in turn is more primitive than †Tharsis and members of the †Varasichthyidae; the †Ichthyodectiformes are more advanced than †Ascalabos. The †Ichthyodectiformes are the plesiomorphic sister-group of the elopomorphs plus the osteoglossomorphs and more advanced teleosts. The osteoglossomorphs are the sister-group of two main clupeocephalan clades: one composed of the clupeomorphs and the ostariophysans and another of the "salmoniforms". The monophyly of Teleostei is supported by one uniquely derived character (quadrate bearing an elongated posteroventral process) and numerous homoplastic characters (e.g., mobile premaxilla, pectoral propterygium fused with first pectoral ray, uroneurals).

The sister-group relationship of Teleostei remains unknown because †Dapedium, †Pachycormidae, †Mesturus, †Aspidorhynchidae, Amia, and [Lepisosteus + †Obaichthys] change their positions depending of the selection of the outgroups. The present study demonstrates that the phylogenetic hypotheses of Teleostei can be changed drastically by changing the members of the outgroup.

Following the new phylogenetic hypothesis, the supercohorts Elopomorpha and Osteoglossocephala, the subcohort Ostarioclupeomorpha, the cohort Clupeocephala, and subcohort Euteleostei sensu ARRATIA (1997), are redefined in a more complete set of characters. These taxa are defined at the primitive level by few uniquely derived characters and several homoplasies. This reflects the problem that teleostean phylogenies face when data from both fossil and extant taxa are considered in phylogenetic analyses.

The addition of new outgroup members (e.g., †Watsonulus, the lepisosteid †Obaichthys, the pycnodont †Mesturus) as well as new characters and new interpretations of certain characters changes previous hypotheses of relationships of teleosts. The inclusion of basal teleosts in an analysis of interrelationships changes the polarity of characters; consequently, the fossil basal teleosts influence the sister-group relationships of extant groups.

J. D. STEWART: A new genus of Saurodontidae (Teleostei: †Ichthyodectiformes) from Upper Cretaceous rocks of the Western Interior of North America
[pp. 335-360, 13 figs., 2 tabs.]

To date, only two genera of saurodontid teleosts have been recognized: Saurocephalus HARLAN 1824 and Saurodon HAYS 1830. LOOMIS (1900) named a new species of Saurodon (Saurodon pygmaeus), based upon a mandible from the Niobrara Chalk Formation of Kansas, U.S.A. Specimens of this species have been published incorrectly as Ichthyodectes ctenodon and Saurodon leanus. The species is valid and cannot be assigned to either of those genera. It exhibits several synapomorphies of the Saurodontidae, but does not share some synapomorphies that unite Saurodon and Saurocephalus; it is a sister-group to those two genera. I therefore propose the generic name Prosaurodon, whose type species is Prosaurodon pygmaeus (LOOMIS 1900). Prosaurodon is known from Santonian and Campanian horizons in North America, but the divergence time for that lineage must be no younger than the Cenomanian, because there are Cenomanian specimens of Saurodonfrom Texas.

The new genus exhibits, in addition to the predentary, three character states that have at times been considered synapomorphies of the Ichthyodectidae: unpaired (fused) parietals; parietal participating in the supraoccipital crest, and basal sclerotics with a serrated margin. Furthermore, the only complete saurodontid skeleton known to date shows a short, remote anal fin which also has been suggested as a synapomorphy of the Ichthyodectidae. New evidence supports the separation of the Saurodontidae plus Ichthyodectidae from the Cladocyclidae. However, the Ichthyodectidae may be paraphyletic with respect to the Saurodontidae.

Saurodontids retained a narrow gular. The presence or absence of this element should be determined for other ichthyodectoid genera.

Francisco José POYATO-ARIZA: The elopiform fish †Anaethalion angustus restored, with comments on individual variation
[pp. 361-368, 4 figs., 1 tab.]

The genus Anaethalion, from the Malm of Bavaria and France, is an elopiform fish of great importance for understanding the origin and early diversification of major teleostean lineages. This paper deals with the type species Anaethalion angustus, which has been previously described in detail, but whose skull and body restorations were incomplete or unknown. These are provided here. The individual variation observed in the caudal skeleton of A. angustus is acknowledged. It involves the number of principal caudal fin-rays and the number of neural and haemal arches of preural centra 2 and 3.

LI Guo-Qing & Mark V. H. WILSON: Early divergence of Hiodontiformes sensu stricto in East Asia and phylogeny of some Late Mesozoic teleosts from China
[pp. 369-384, 7 figs., 1 tab., 2 apps.]

Hiodontoidea as constituted by GREENWOOD in 1970, consisting of †Lycopteridae and Hiodontidae, are paraphyletic. Recent phylogenetic studies of Osteoglossomorpha suggest instead a monophyletic Hiodontiformes sensu stricto, sister to Osteoglossiformes and with a minimum age range of Early Cretaceous to Recent. †Lycopteridae are not hiodontiforms but are stem-group osteoglossomorphs, sister to all extant clades of Osteoglossomorpha. Since 1976, numerous taxa named on materials from the Upper Mesozoic of East Asia have been added to the †Lycopteridae and Hiodontidae and were included in the Hiodontoidea GREENWOOD 1970, but without any formal phylogenetic analysis (e.g., †Paralycoptera, †Yanbiania, †Changichthys, and †Jiuquanichthys). This study confirms that †Yanbiania is the earliest representative of the Hiodontiformes yet known, and that the five hiodontiform synapomorphies (dermosphenotic triradiate, opercle an irregular parallelogram, hyomandibula double-headed, dorsal arm of posttemporal more than twice as long as ventral arm, and seven-rayed pelvic fin) were already present in †Yanbiania. The 'Mid'-Cretaceous †Plesiolycoptera is considered to be a stem-group hiodontiform because of the presence of a triradiate dermosphenotic, a posttemporal with a dorsal arm at least twice as long as the ventral arm, and a seven-rayed pelvic fin. However, our study does not support inclusion of either †Jiuquanichthys or †Changichthys in the Hiodontiformes because neither shares the five hiodontiform synapomorphies. †Jiuquanichthys is closely related to †Kuyangichthys and †Tongxinichthys and together these three genera form a more primitive stem-group osteoglossomorph clade that is sister to †Lycopteridae plus all other Osteoglossomorpha. In addition, a possible sister-group relationship between †Huashiidae and Heterotidinae is suggested. †Tanolepis and †Paralycoptera are sister taxa forming a clade that is sister to all other members of the Osteoglossoidei. The fossil record of East Asia thus includes stem-group members of the Osteoglossomorpha, Hiodontiformes, Osteoglossoidei, and Heterotidinae.

ZHANG Jiang-yong & JIN Fan: A revision of †Tongxinichthys MA 1980 (Teleostei: Osteoglossomorpha) from the Lower Cretaceous of northern China
[pp. 385-396, 6 figs., 1 tab., 1 app.]

†Tongxinichthys is reviewed based mainly on specimens from Madongshan Formation of Liupanshan Group, Tongxin, Ningxia of China. The osteology is redescribed and bones such as the supraoccipital, exoccipital, prootic, urohyal, scapula, coracoid, and mesocoracoid are described for the first time.

A cladistic analysis of the phylogenetic relationships among seven Chinese fossil osteoglossomorph genera (†Kuyangichthys, †Jinanichthys, †Jiuquanichthys, †Lycoptera, †Paralycoptera, †Tongxinichthys, and †Yanbiania) and three living osteoglossomorphs (Hiodon, Notopterus, and Osteoglossum) suggests that †Tongxinichthys is a sister-genus of †Jiuquanichthys. The sister-group relationship [†Tongxinichthys + †Jiuquanichthys] is supported by one uniquely derived character (supraoccipital with a long, narrow process on its anterior margin). In the strict consensus tree, [†Tongxinichthys + †Jiuquanichthys], †Lycoptera, the clade [†Kuyangichthys + [†Jinanichthys + [†Paralycoptera + [Notopterus + Osteoglossum]]]], and [Hiodon + †Yanbiania] form a polytomy. Therefore, their phylogenetic relationships are still unresolved.

Alison M. MURRAY & Mark V. H. WILSON: Contributions of fossils to the phylogenetic relationships of the percopsiform fishes (Teleostei: Paracanthopterygii): order restored
[pp. 397-411, 2 figs., 1 tab.]

The contribution to phylogeny by fossil members of the order Percopsiformes, a North American teleostean group of freshwater fishes, is assessed in a study of basal paracanthopterygians based on osteological characters. In 1989, PATTERSON & ROSEN published a cladistic study suggesting that the Percopsiformes were a paraphyletic assemblage. In recent years, several new taxa of basal paracanthopterygians have been described from the fossil record, including the freshwater Palaeocene 'percopsiform' genera †McconichthysGRANDE 1988, †Massamorichthys MURRAY 1996, and †Lateopisciculus MURRAY & WILSON 1996, and the marine Early Cenomanian sphenocephalid genus †Xenyllion WILSON & MURRAY 1996.

Cladistic analysis of 46 osteological characters (including 19 of the 20 used by PATTERSON & ROSEN 1989) indicates that the order †Sphenocephaliformes, consisting of †Sphenocephalus and †Xenyllion, is the primitive sister to the rest of the Paracanthopterygii. Inclusion of the extant Amblyopsidae within the Percopsiformes greatly increases the tree length. Instead, the Amblyopsidae are considered to lie within the Anacanthini, although relationships within Anacanthini have not been fully determined by the present study. The order Percopsiformes is monophyletic after the removal of the Amblyopsidae. Percopsiform synapomorphies include loss of the "gadoid notch" from the postmaxillary process of the premaxilla, fewer than seven branchiostegal rays, fusion of the parhypural with the lower hypural plate, and an opercle dorsal projection that is truncated anteriorly. A majority-rule consensus tree suggests that †Mcconichthys and †Libotonius are stem percopsiforms, whereas †Amphiplaga, †Erismatopterus, †Lateopisciculus, and †Massamorichthys are more closely related to Percopsis than they are to Aphredoderus.

Lance GRANDE & William E. BEMIS: Historical biogeography and historical paleoecology of Amiidae and other halecomorph fishes
[pp. 413-424, 7 figs.]

The historical biogeography of the family Amiidae is reviewed, and cast in a vicariance model, based on much recently published phylogenetic information on the group. We believe that vicariance studies are a logical first step in any rigorous historical biogeographic study. Consequently, a hypothesis of relationships based on non-biogeographical data is also a necessary initial step in a thorough study of historical biogeography. Recently we published a comprehensive phylogenetic study of amiid fishes, which divided the family into four subfamilies: Amiinae, †Vidalamiinae, †Solnhofenamiinae, and †Amiopsinae. Among these subfamilies, †Vidalamiinae show the strongest vicariant patterns.

The historical paleoecology of the family (focussing on freshwater vs. marine origins for taxa) is also explored within a phylogenetic framework. Although the extant Amia calva (and apparently the subfamily Amiinae) is freshwater, the family Amiidae appears to primitively have been marine.

Terry GRANDE: Distribution patterns and historical biogeography of gonorynchiform fishes (Teleostei: Ostariophysi)
[pp. 425-444, 13 figs.]

Using the phylogenetic analysis of T. GRANDE & POYATO-ARIZA (1999) as the basis, the biogeography of gonorynchiform fishes are investigated. The biogeographic patterns observed are then analyzed for historical significance using cladistic vicariance techniques. Results of this study show that the history of Gonorynchiformes is complex. The distribution patterns observed today most likely result from a combination of dispersal and vicariant events. Some groups (e.g., †Dastilbe and †Parachanos) show the classic transatlantic pattern between eastern Brazil and Africa. Other groups have a distribution pattern that indicate a possible Tethys origin. The Gonorynchidae (e.g., †Notogoneus) shows great potential for future biogeographic studies, but its phylogenetic relationships must be resolved further before its area relationships can be interpreted. Previous published dispersal scenarios concerning gonorynchid biogeography are based solely on the fossil record, and need to be tested with better resolved phylogenetic patterns and new taxa currently being studied.

Eugenia K. SYTCHEVSKAYA: Freshwater fish fauna from the Triassic of Northen Asia
[pp. 445-468, 17 figs.]

The two principal freshwater fish faunas known from the Triassic of Northern Asia have been accounted and/or re-examined. Among the earliest Triassic assemblage of the Tunguska River Basin (Siberia), the presence of the perleidid Eoperleidus, formerly cited as nomen nudum, has been confirmed. Other genera such as the scanilepiform Evenkia, the semionotiform Tungusichthys, and the pholidopleurid Arctosomus are redescribed. "Tungusichthys" of SELEZNEVA (1988) is shown to be based on the type material of Eoperleidus.

From the Ladinian-Carnian Madygen assemblage, Fergana Depression (Kirghizstan), six actinopterygian genera are recognized, five of them new. These include the evenkiid Oshia, the palaeoniscids Ferganiscus and Sixtelia, the perleidids Megaperleidus and Alvinia, and a new species of Saurichthys. The presence of some shared endemic groups in the two surveyed Asiatic assemblages, primarily the evenkiids, supports the idea that the Triassic vertebrate biota of Northern Asia belongs to a specific biogeographic province.

CHANG Mee-mann: 'Mid'-Cretaceous fish faunas from northeast China
[pp. 469-480, 7 figs., 3 tabs.]

The 'mid'-Cretaceous fish faunas from northeast China are distributed mainly in Songliao Basin, Heilongjiang and Jilin Provinces and in a series of small basins that extend east to Songliao Basin. In contrast to the early Early Cretaceous endemic †Lycoptera-fauna from North China the members of the 'mid'-Cretaceous (approximately Aptian-Cenomanian) fish faunas from northeast China find their contemporary relatives either in neighboring areas or even in regions long distance away. Included in these faunas are species belonging to Osteoglossomorpha and Euteleostei. Those from the late Early Cretaceous are osteoglossomorphs (†Plesiolycoptera daqingensis ZHANG & ZHOU 1976, †Jiaohichthys pulchellus MA 1983, †Pulinia baojiatunensisMA 1983 and †Yanbiania wangqingica LI 1987) and primitive euteleosts (†Manchurichthys uwatokoi SAITO 1936, preliminarily referred to Salmoniformes incertae sedis). Those from the Late Cretaceous fish fauna are primitive euteleosts (†Sungarichthys longicephalus TAKAI 1943 and †Jilingichthys rapax ZHOU 1976, both preliminarily referred to Salmoniformes incertae sedis), a neoteleost (†Hama macrostoma ZHOU 1976), and a single tooth of Selachii. The osteoglossomorphs of approximately the same age are also found in the adjacent areas of North America. The primitive euteleosts and neoteleosts seem to have wider distribution, possibly related to the Late Cretaceous general transgression in the northern hemisphere. Comparison of †Jiaohichthysand †Yanbiania shows surprising morphological similarity between the two genera and the necessity of inclusion of †Jiaohichthys in further phylogenetic analyses of osteoglossomorphs. The stratigraphical distribution of the fishes shows different combinations in different formations. The lower (older) formations contain osteoglossomorphs and basal euteleosts, while the higher one contains only osteoglossomorphs and the highest one – basal euteleosts, a neoteleost, and a selachian tooth. The successive change of some groups of fishes in the section is obviously age-related while the presence or absence of certain groups of fishes in certain formations is suggested to be related to different environments in which the fishes occur.

Toni BÜRGIN: Middle Triassic marine fish faunas from Switzerland
[pp. 481-494, 8 figs., 2 apps.]

At Monte San Giorgio (Southern Switzerland) over the last 85 years large quantities of chondrichthyans and especially of actinopterygian fishes have been found in marine Middle Triassic sediments. Most of these fossils are from the Anisian-Ladinian boundary zone and were deposited in the anoxic bottom layers of a shallow lagoon. Other contemporaneous fossil sites in Eastern Switzerland have been systematically studied only over the last years. Their increasing amounts of fossil actinopterygians show close similarities with some of the Monte San Giorgio fossiliferous beds. The present status of these studies is reported. A comparison with other Middle Triassic fish faunas from Europe shows either close relationships (Alpine realm) or certain dissimilarities (Muschelkalk) which can be explained by different paleoecological and paleogeographical conditions.

Andrea TINTORI & Cristina LOMBARDO: Late Ladinian fish faunas from Lombardy (North Italy): stratigraphy and paleobiology
[pp. 495-504, 6 figs., 1 tab.]

The Ladinian (Middle Triassic) turns out to be the richest period for Mesozoic fishes. Besides well known faunas, such as that from the early Ladinian of the Besano-Monte S.Giorgio Grenzbitumenzone, new fossiliferous levels have been exploited in the last years. One of these levels, the latest Ladinian Kalkschieferzone, yielded a bounty of fossil fishes, that were collected both at Ca' del Frate (Viggiù, Italy) and Meride (Canton Ticino, Switzerland) fossil sites. This latter large collection of well preserved specimens allows detailed paleobiological studies: a restoration of the ontogenesis and the identification of sexual dimorphism, e.g., were achieved. All the well represented species of fishes show year-class sampling, with a sequence of character states (squamation becoming complete, body and fin shape modifications) which are interpreted as due to the growth process, herein, and not considered to represent different species in the same genus. Adult specimens of some of these same species are also known from the area of Perledo, where, in the last century, a few dozens of fishes have been found during quarry works. The Perledo fauna is only partially contemporary with the one of the Kalkschieferzone, yielding also lower Ladinian species. The different composition of both respective Late Ladinian fish faunas, together with sedimentology, are best explained by different paleoenvironments. At Perledo only adults are found, which indicates a more open sea, whereas at Ca' del Frate – Meride juveniles are common, suggesting that this more protected area was used for reproduction and nursery.

Francisco José POYATO-ARIZA, Ángela D. BUSCALIONI & Joan CARTANYÀ: The Mesozoic record of osteichthyan fishes from Spain
[pp. 505-533, 7 figs., 2 tabs., 1 app.]

The record of Mesozoic osteichthyan fishes from Spain is evaluated: the faunas of 51 localities are revised and catalogued, updating their age. This record is analysed from an integrative multidisciplinary perspective.

The Mesozoic localities of Spain are mostly restricted to the north and central eastern half of the Iberian Peninsula, being most of them Cretaceous in age. The osteichthyan record, as retrieved from literature, comprises 50 genera grouped into 29 families, together with 25 incertae sedis or undetermined taxa. The record is patchy; 53 % of the entire generic diversity is provided by three fossil-Lagerstätten only (Mont-ral, Middle Triassic; Las Hoyas and El Montsec, both Early Cretaceous), which have yielded abundant complete and articulated remains. The other 48 localities have yielded only relatively diverse isolated remains. The most extensively cited actinopterygian genus is Lepidotes, followed by Coelodus and other pycnodonts, most of them being unreliably identified on the basis of isolate remains. The Sarcopterygii are represented by a scarce record of the family Coelacanthidae only.

The metric completeness analysis at family level and series-stages reveals that the Mesozoic fossil record of osteichthyans from Spain presents a 54 % of completeness. Three major groups of Actinopterygii have been analysed to approach their completeness: 1) the non-halecostome Actinopterygii are restricted to the Triassic, associated to the Mont-ral fauna; 2) the record of non-teleostean Halecostomi is affected by the poorly informative beds from the Jurassic, thus showing the lowest values of completeness: 48 %; against 100 % of non-halecostome Actinopterygii and 70 % of Teleostei. The non-teleostean Halecostomi include the most extended Lazarus taxa; 3) finally, the Teleostei families are distributed throughout the Early Cretaceous and Late Cretaceous, with an unexpected lack of record in the Aptian-Albian.

The worst documented records correspond to the Jurassic stages, in which the number of localities, the diversity, and the completeness of the stratigraphic record are quite poor.

Joan CARTANYÀ: An overview of the Middle Triassic actinopterygians from Alcover, Mont-ral and El Pinetell (Catalonia, Spain)
[pp. 535-551, 13 figs.]

The fossils from the Alcover dolomites have been known since 1963. The laminated Alcover dolomite Unit ("Pedra d'Alcover") is up to 70-80 m thick, and fills the interreefal depressions. Stratigraphically, it is placed in the upper Ladinian (Middle Triassic). Anoxic and hypersaline conditions were developed periodically during the deposition of the Alcover Unit. The main facies are massive dolmicritic beds (1-10 cm thick) and thin-bedded laminated dolmicrites. Locally convoluted beds with intraformational folds and slides are present.

The fossil assemblage is allochthonous and composed of land plants, jellyfishes, brachiopods, molluscs, arthropods, echinoderms, fishes, and reptiles. The reputation of these localities is based on the fact that the fossils are preserved in dolmicrite and of their ichthyofaunal diversity. The Alcover fish fauna is characterized by predominantly medium-sized specimens. Saurichthyiformes and Perleidiformes are the most common actinopterygians in the beds. At present, the better preserved fossil actinopterygians are classified at generic level, and in some cases to family level only. Altogether, 15 genera belonging to 11 families and nine orders have been identified. These are: Ptycholepis, ?Boreosomus, Saurichthys, Colobodus, Perleidus, ?Ctenognatichthys, Peltoperleidus, ?Cleithrolepididae indet., Peltopleurus, Peripeltopleurus, ?Platysiagum, Luganoia, Eosemionotus, Archaeosemionotus, Allolepidotus, Eoeugnathus, ?Caturus, ?Ophiopsis, and ?Pholidophoridae indet.

Didier B. DUTHEIL: An overview of the freshwater fish fauna from the Kem Kem beds (Late Cretaceous: Cenomanian) of southeastern Morocco
[pp. 553-563, 8 figs., 2 tabs.]

A new fish fauna from the Upper Cretaceous continental formation of southeastern Morocco is reported. Sieving and gleaning in detritic beds of this formation has provided thousands of disarticulated bones, scales and teeth. One locality provided articulated elasmobranchs and actinopterygians. Articulated fossil bichirs were found therein for the first time with at least five undescribed teleosts. The structural organization of the muscles in most of these fishes are in high quality of preservation. The habitat suggested for this fish fauna is a lake or a pool deposit.

Gloria ARRATIA & Hans-Peter SCHULTZE: Mesozoic fishes from Chile
[pp. 565-593, 21 figs., 1 app.]

Fossil fishes are known from few Chilean localities ranging from the Paleozoic to the Quaternary. Among these, the Jurassic localities of the Cordillera de Domeyko in northern Chile are outstanding because of the presence of basal teleosts among a diversified fish fauna. In addition, preservation is of special quality including soft tissue structures.

Perleidiforms are reported for the first time from the Triassic of Chile (Quebrada San Pedrito). Recently, numerous small proleptolepid teleosts, belonging to a new genus, were recovered from Sinemurian localities of Quebrada La Carreta, as well as a new coelacanth from a Sinemurian locality at Vaquillas Altas (the first record from Chile). The proleptolepids represent at least two forms. They differ from the European proleptolepids in size of the maxilla and jaw dentition. They add important new information on the neurocranium, hyoid arch, and pectoral girdle. Several teleostean species and a few halecostomes are known from the Oxfordian of Cerritos Bayos, Quebrada del Profeta, and Sandón. Most teleosts represent endemic groups. The pachycormiforms, pycnodontiforms, and semionotiforms have not been studied in detail but apparently correspond to taxa that have a cosmopolitan distribution. The earliest record of a pachyrhizodontoid fish is reported from the Upper Jurassic locality Termas del Flaco in central Chile. The first record of aspidorhynchiforms in Chile is reported from the Lower Cretaceous of Lomas Negras near Calama in northern Chile; differences in the predentary and opercular bones do not permit us to assign this material to any known genus.

Alexander MUDROCH, Detlev THIES & Albrecht BAUMANN: ⁸⁷Sr/⁸⁶Sr analysis on Late Jurassic fish teeth. Implications for paleosalinity of fossil habitats
[pp. 595-604, 7 figs., 1 app.]

The study by SCHMITZ et al. (1991) of ⁸⁷Sr/⁸⁶Sr ratios in various Phanerozoic fishes led to the idea of testing this method on fish faunas from the Upper Jurassic of Germany and France. These faunas consist of microvertebrate remains. The Sr isotope results for the Kimmeridgian data from northern Germany are consistent with faunistic evidence that implies an environment of lower marine salinity. The high radiogenic content of the material from the Boulonnais, northern France (Kimmeridgian to Tithonian) contradicts the faunistic implications that point more probably to a genuine marine environment for the sampled periods. The possibilities of postmortem alteration of the measured ⁸⁷Sr/⁸⁶Sr ratios are discussed.

The Mesozoic era was an important time in the evolution of chondrichthyan and actinopterygian fishes because it was then that most of the modern groups first entered the fossil record and began to radiate. By the end of the era, many archaic forms had disappeared and the foundation had been laid for the ichthyofauna that now exists. Despite this significant evolutionary change, before 1990 there had been little concerted research done on Mesozoic fishes and no synopsis or compilation of the systematics and paleoecology of Mesozoic fishes had been published, not even for single groups. To remedy this deficiency, Gloria ARRATIA initiated the symposium "Mesozoic Fishes". The first meeting "Mesozoic Fishes - Systematics and Paleoichthyology" was held in Eichstätt from August 9 to 12, 1993 and the first volume of Mesozoic Fishes, including 36 papers concerning elasmobranchs, actinopterygians and sarcopterygians and the paleoecology of certain important fossil localities was published in 1996.

Gloria ARRATIA and Hans-Peter SCHULTZE organized the second Symposium with the help of the Institut für Paläontologie, Museum für Naturkunde, Berlin, and the Deutsche Forschungsgemeinschaft (German Science Foundation), Bonn, Germany. It was held in Buckow from July 6 to 10, 1997. Buckow is a small village about 45km east of Berlin, surrounded by lakes and forests, which offered a quiet and pleasant environment for the meeting.

The participants were able to stay together in a residence connected to the meeting hall, an arrangement that created a friendly and familial atmosphere. The oral presentations during the day were followed by informal talks on some special fossil localities after dinner. After the poster presentation, participants were able to examine material and discuss their ideas, often with specimens in hand.

The group of Mesozoic fish researchers is small; nevertheless, the members of the group are enthusiastic and hard-working. Many of the participants gave more than one presentation. The wide-ranging interests of the group are reflected in the content of this volume. The papers provide new information that helps to clarify the course of evolution of certain chondrichthyans and many actinopterygians in the Mesozoic and in more recent times.

The results of the symposium presented in this volume reflect the current state of knowledge of Mesozoic fishes. Phylogenetic relationships of chondrichthyans and actinopterygians are the central issue. In addition, attention is given to questions of morphology and to the Mesozoic fossil record in a variety of countries such as southern Asia, Chile, China, Italy, Morocco, Spain, and Switzerland. The new findings described in the 31 papers and the disagreements among authors concerning interpretations of characters and phylogenetic relationships of actinopterygian subgroups are an exciting invitation to further research.

Gloria ARRATIA and Hans-Peter SCHULTZE