A new lepisosteiform neopterygian (Actinopterygii) from the lower Toarcian Grimmen Formation (Lower Jurassic) of Mecklenburg and Western Pomerania, Germany

Abstract

The Lower Jurassic (lower Toarcian) Grimmen Formation (formerly known as "Green Series") of Mecklenburg and Western Pomerania (Northeast Germany) yielded two articulated anterior skeletons of fossil fish, which represent a new lepisosteiform genus and species: Mengius acutidens gen. et sp. nov. The new taxon is described. Osteology of the species is most similar to the contemporaneous species of Lepidotes Agassiz. It differs from Lepidotes in having a non-tritoral dentition and a differently shaped operculum. A cladistic analysis shows Mengius acutidens gen. et sp. nov. to be the sister taxon of Lepidotes, accordingly, the new species is assigned to the family Lepidotidae Agassiz. The occurrence of M. acutidens emphasises the endemic character of the lower Toarcian Grimmen Formation fish fauna of Northeast Germany.

Introduction

Agassiz (1832) introduced the actinopterygian genus Lepidotes for certain deeply fusiform fossil fishes from the Posidonia Shale Formation (Lower Jurassic, lower Toarcian) of Baden-Württemberg in South Germany. Later, Agassiz (1833, 1834, 1837) changed the spelling of the taxon into Lepidotus which then became widely used by scholars worldwide. Modern authors (e.g. Wenz 1968, Thies 1989a) have, however, returned to the original spelling Lepidotes.

The long taxonomic history of Lepidotes and the species referred to this genus is complex. According to Agassiz (1832), Lepidotes gigas Agassiz, 1832 is to be considered as the type species of the genus. Taxonomical confusion arose from Agassiz (1833, 1837) and Quenstedt (1852, 1858, 1847) who both considered L. gigas as conspecific with Lepidotes elvensis (Blainville, 1818) from the Lower Jurassic of France. This point of view was accepted by Woodward (1895) who listed L. gigas as a younger synonym of L. elvensis. Further, Woodward (1895) gave a comprehensive diagnosis of the genus and elaborated an identification key for the numerous species of Lepidotes already known at that time.

Jaekel (1929), in a study of the genera Lepidotes and Leptolepis from the Lower Jurassic of Mecklenburg (Northeast Germany), doubted that L. gigas from Baden-Württemberg and L. elvensis from North France were conspecific. Instead, he proposed the new name of Lepidotes quenstedti for the specimens from the Lower Jurassic of South Germany, however, without presenting a systematic definition (diagnosis and description) or designating a specific specimen as the holotype. L. quenstedti should therefore be considered as a nomen nudum.

Recently, one of us (D.T.) searched for the holotype of “Lepidotus”buelowianus Jaekel, 1929 from the Lower Jurassic of Dobbertin (Mecklenburg). Jaekel based this taxon on a single specimen that he briefly described. He added to the description a schematic black and white line drawing of the skull but did not include any photographs (Fig. 1). A photograph of the specimen was published only later by Kurd von Bülow (1952: fig. 4) showing that Jaekel’s drawing is quite accurate with regard to preservation and orientation of the fossil.

Fig. 1

Reproduction of Jaekel’s (1929) fig. 1 showing his line drawing of the cephalic skeleton of the holotype and single specimen of "Lepidotus" buelowianus from the Lower Jurassic of Dobbertin, Mecklenburg, Northeast Germany. The specimen is lost

The specimen of “L.” buelowianus was given to Jaekel by his former assistant Ernst Ulrich von Bülow-Trummer at the beginning of the 1920s. At that time, Jaekel had the position of professor of geology at the University of Greifswald. He retired in 1928 and went to China where he died of pneumonia in Beijing in 1929. In Jaekel (1929: 14) is stated: "Da der Autor vor der Drucklegung nach China verreiste, konnte er die Korrektur nicht selbst vornehmen." [The author went to China before printing and therefore could not do proofreading himself]. This indicates that he finished the manuscript on “L.” buelowianus early in 1928 or before. It may be assumed that the type and only specimen was deposited in the collection of the Institute of Geography and Geology of the University of Greifswald. Actually, the collection houses a specimen of Lepidotes (no cat. no.) labelled as the holotype of “L.” buelowianus. The specimen in question consists of a skull and partial anterior trunk and belongs undoubtedly to Lepidotes but is in a very bad state of preservation. Regardless, it is obvious that it differs considerably from the specimen described and figured by Jaekel (1929). Most importantly, it displays the right side of the body whereas Jaekel’s specimen of “L.” buelowianus shows the left side. The uncatalogued specimen in the Greifswald university collection does therefore not represent the holotype of “L.” buelowianus. Another specimen of Lepidotes agreeing morphologically with Jaekel’s (1929) description and figure was not found in the collection. So, the true holotype of “L.” buelowianus is not kept in the Greifswald university collection and appears to be lost. Also, Jaekel’s (1929) description of “L.” buelowianus lacks diagnosis and proper measurements. At present, it may therefore be best to consider “L.” buelowianus as a nomen dubium as well.

Apart from the uncatalogued specimen erroneously labelled as holotype of “L.” buelowianus the collection of the Greifswald Institute of Geography and Geology holds further remains of Lepidotes from the lower Toarcian of Mecklenburg and Western Pomerania. Amongst these are two specimens (cat. nos GG 435 (old number: 4/118) and GG 23093) that each consist of the skull and anterior trunk. Both were already studied and published previously. In specimen GG 435, the cranial sensory line system includes a pit line on the maxilla, which is only rarely seen in fossil actinopterygians (Thies 1989b). In specimen GG 23093 the gut contents are visible containing shell fragments of small crustaceans (Thies et al. 2019). A re-investigation of these two specimens has now revealed that they obviously do not belong to any known species of Lepidotes from the Lower Jurassic (lower Toarcian) of Europe. We consider them as a new taxon which is described in this study.

Geological setting

In Northeast Germany, the lower Toarcian succession includes the Grimmen Formation, formerly known as “Green Series” (Grüne Serie), which is the time equivalent of the Posidonia Shale Formation of southern and western Germany (Ansorge et al. 2024). The Grimmen Formation comprises some tens of metres of greyish-green clay of the falciferum Zone, overlying clastic sediments of the semicelatum Subzone (upper part of the tenuicostatum Zone) and brown clay of the elegantulum Subzone (Ansorge et al. 2024).

In Mecklenburg and Western Pomerania, Toarcian marine sediments crop out only in the abandoned open-cast clay pits of Grimmen and Dobbertin, both exposing a succession that was glacially dislocated during the Pleistocene ice advances (see Ernst 1967, 1991, 1992; Ansorge and Obst 2007, 2015; Ansorge and Grimmberger 2016; Ansorge et al. 2024). The basal part of the Grimmen Formation (Reinberg Member) clay was deposited during the Toarcian Oceanic Anoxic Event (ca. 181.5 Ma; Jenkyns 1988, Ogg and Hinov 2012) washed by rivers draining Fennoscandia.

In the Reinberg Member, rather small light grey laminated carbonate concretions or nodules of the exaratum Subzone occur in which fossils are extraordinarily well preserved, sometimes three-dimensionally (Kirejtshuk and Ansorge 2022). These syn- or early diagenetic micritic concretions of about 2–3 cm thickness formed under anoxic calm water conditions without bioturbation of the sediment.

The size of the fish fossils is restricted by the size of the concretions which are larger in Dobbertin than in Grimmen. Articulated fishes are usually dorsoventrally embedded. The upper side is often compressed and sometimes distorted. Isolated bones are much more common than articulated skeletons. Isolated bones together with ammonite and Inoceramus shell debris are often enriched in regurgitalites of larger vertebrates.

Material and methods

Material

The material consists of two incomplete but articulated fish specimens from the Lower Jurassic (lower Toarcian) of Mecklenburg and Western Pomerania (Northeast Germany). Both specimens are housed in the geological collection of the Institute of Geography and Geology of the University of Greifswald under the catalogue numbers GG 435 and GG 23093.

For comparison of scale histology, an isolated scale was taken from an adult specimen of Lepidotes sp. from the Lower Jurassic (lower Toarcian) Posidonia Shale Formation of Schandelah east of the city of Braunschweig, Northwest Germany. The specimen is housed in the geological and paleontological collection of the Geopark-Trägerverein Braunschweiger Land–Ostfalen e.V. (cat. no. IZ-74–1). The scale is deposited in the geological collection of the Institute of Geography and Geology of the University of Greifswald (cat. no. GG 435/1). Details of the palaeontology and bio- and lithostratigraphy of the lower Toarcian of the Braunschweig area are given by e.g. Hauff et al. (2014) and Mutterlose et al. (2022).

Preparation

GG 435: The method of preparation is unknown. The appearance of the specimen showing many delicate details indicates that preparation methods included acid preparation by which most of the carbonate matrix on the left side of the body was removed. The right side is still covered by matrix. The sort of acid, its concentration, sort of solution buffering and duration of treatment are, however, unknown. Two scales of this specimen were removed for preparation of histological thin-sections with cover glasses. Lateral sections of the scales were fixed on glass slides in epoxy resin and ground down to a thickness of ca. 30 µm (GG 435/1).

GG 23093—This specimen was prepared mechanically by J. A. using a vibrating engraver (Vibrograv^®). The specimens were dusted with ammonium chloride (NH₄Cl) before photographing.

Osteological description

In the morphological description of the cephalic skeleton, we use the nomenclature for cranial ossifications proposed by Schultze (2008), which at present is the best expression of homology between individual bones of actinopterygians and sarcopterygians within Osteichthyes. The terminology for the description of the sensory line system follows Devillers (1958) and Thies (1989a, b; 1991).

Cladistic analysis

To infer the phylogenetic position of the new taxon, a cladistic analysis was performed. Character states of the new taxon were scored based on the character matrix of López-Arbarello and Sferco (2018). The analysis was performed with TNT 1.5 (Goloboff and Catalano 2016). The analysis was performed with the “traditional search” option using standard parameters, except setting the number of replications to 1.000. Following suggestions of Goloboff et al. (2018) for morphological data-sets, implied weighting was applied with a constant of concavity (k) of 12. To test for tree stability, we employed the jackknifing resampling option of TNT with a deletion probability set to 50% (following suggestion by Goloboff et al. 2003) and 10.000 replications.

Morphological and anatomical abbreviations

Ang—angular, apl—anterior pit line of the skull roof, Art—articular, Ch—ceratohyal, Cl—cleithrum, Co—coronoid, coio—canal for the otic section of the infraorbital sensory line, Dpal—dermopalatine, Dsp—dermosphenotic, Dspl—dentalosplenial, Ectp—ectopterygoid, Entp—entopterygoid, Ex—extrascapular, Hh—hypohyal, hpl—horizontal pit line of the cheek, Io—infraorbital, Iop—interoperculum, La—lacrimal, mdpl—mandidular pit line, mpl—median pit line of the skull roof, Mtp—metapterygoid, Mx—maxilla, N I—foramen for the olfactory nerve, Na—nasal, Op—operculum, Pa—parietal, pccll—pores of the canal for the cephalic lateral line, pcio—pores of the canal for the infraorbital sensory line, Pcl—postcleithrum, pcoio—pores of the canal for the otic section of the infraorbital sensory line, pcpocll—pores of the canal for the postotic section of the cephalic lateral line, pcsc—pores of the canal for the supratemporal commissure, pcsop—pores of the canal for the supraorbital sensory line, pmdc—pores of the mandibular canal, pmpl—pores of the maxillary pit line, Pmx—premaxilla, poc—pores of the oral canal, Pop—preoperculum, Ppa—Postparietal, ppc—pores of the preopercular canal, ppl—posterior pit line of the skull roof, Pt—posttemporal, Qu – quadrate, Quj—quadratojugal, Rb—branchiostegal rays, r.pal.VII—foramen for the terminal branch of the palatine nerve, Sang—surangular, Scl—supracleithrum, Scb—sclerotic bones, Smx—supramaxilla, So—supraorbital, Sob—suborbital, Sop—suboperculum, Spo—sphenotic, Stta – supratemporotabular, Sym—symplectic, Vo—vomer, (l) —left, (r) —right, ? —questionable identification.

Institutional abbreviations

GG—Institut für Geographie und Geologie, Universität Greifswald.

Systematic palaeontology

Class Osteichthyes Huxley, 1880

Subclass Actinopterygii Cope, 1887

Series Neopterygii Regan, 1923

Superdivision Holostei Müller, 1846 sensu Grande 2010

Division Ginglymodi Cope, 1872 sensu Grande and Bemis 1998

[see Remark 1 in Appendix]

Order Lepisosteiformes Hay, 1929

[see Remark 2 in Appendix]

Family Lepidotidae Agassiz, 1833

1860 Lepidotidae—Owen: 143

1860 Stylodontes—Wagner: 81 (in part)

1863 Sphaerodontes—Wagner: 617

1880 Sphaerodontidae—Günther: 368

1890 Semionotidae—Woodward: 30 (in part)

1890 Stylodontidae—Zittel: 202 (in part)

1890 Sphaerodontidae—Zittel: 207 (in part)

[see Remark 3 in Appendix]

Diagnosis. The family comprises medium to large sized fishes of mostly fusiform body shape. Synapomorphic characters include (1) large and knob-like anterior process of posttemporal bone, (2) ganoid scales with serrate or dentate posterior margin, (3) parietals (= "frontals") tapering gradually, and (4) ventral border of infraorbital series flexes abruptly dorsally at the anterior margin of the orbit. (1) is uniquely derived and (2) to (4) are homoplastic.

[see Remark 4 in Appendix]    

Genus Mengius nov.

ZooBank LSID. zoobank.org:act:B43DF887-B6EC-4E80-B2BD-4A1EFCB84D01

Derivation of name. After Dr. Stefan Meng honouring his merits in keeping the geological collection of the Greifswald University Institute of Geography and Geology.

Diagnosis. The genus contains fusiform fishes of moderate size. The scales are coarsely serrated at the posteroventral margin. The following characters are diagnostic: (1) longitudinal articulation of the scales only by a strong anterior dorsal process, (2) relative position of the lower jaw articulation at the level around the centre of the orbit, (3) symplectic is not directly involved in the jaw joint, (4) length to width ratio of parietals (= "frontals") in adult sized individuals equals or is larger than three, (5) preorbital portion of parietal (= "frontal") not significantly large, (6) left and right nasal bones close to each other, (7) most anterior supraorbital bone expanded anteroventrally, making the anterodorsal corner of the orbit, (8) posterior rim of the orbit formed by three infraorbital bones, (9) distance from anterior tip of premaxilla to posterior edge of operculum equals approximately four times maximum diameter of orbit, (10) parietals only slightly tapering anteriorly, (11) one pair of extrascapulars, (12) extrascapulars abut in dorsal midline, (13) external surface of skull roof smooth, (14) number of suborbitals three or four, (15) preoperculum slightly curved, without clear definition of dorsal and ventral arm, (16) ventral margin of operculum convex and confluent with caudal margin, (17) number of postcleithra four, (18) maxillary pit line present, (19) oral canal present, (20) dentition non-tritoral, (21) premaxillary and dentary teeth of piercing type, tall, slender, styliform, pointed.

Characters (1) to (8) define the node in our cladogram for Mengius acutidens. They are taken from López-Arbarello and Sferco (2020), and in definition and state they correspond to their characters 9, 96, 83, 101, 104, 107, 118, and 124. The additional characters (9) to (21) are all primitive or homoplastic but in combination we consider them as derived for Mengius.

Type species. Mengius acutidens gen. et sp. nov.

Mengius acutidens sp. nov.

Figures 2, 3, 4, 5, 6a, 7

Fig. 2

Mengius acutidens gen. et sp. nov. Holotype, specimen GG 435. a left lateral view of the specimen, b enlargement of the snout showing details of the mandibular arc and the dentition. Specimen dusted with NH₄CL for photography

Fig. 3

Mengius acutidens gen. et sp. nov. Holotype, specimen GG 435. a left lateral view of the skull and branchial apparatus, b schematic drawing of the cephalic skeleton (labelling of infraorbital 3 omitted). Specimen dusted with NH₄CL for photography

Fig. 4

Mengius acutidens gen. et sp. nov. Paratype, specimen GG 23093. a left lateral view of the specimen, b right lateral view of the specimen. Specimen dusted with NH₄CL for photography

Fig. 5

Mengius acutidens gen. et sp. nov. Paratype, specimen GG 23093. a left lateral view of the skull and branchial apparatus, b schematic drawing of the cephalic skeleton and adjoining scale rows. Specimen dusted with NH₄CL for photography

Fig. 6

Longitudinal thin-sections of scales. a Mengius acutidens gen. et sp. nov. (holotype, GG 435), b Lepidotes sp. (lower Toarcian Posidonia Shale Formation of Schandelah, Braunschweig area, Northwest Germany. GG 435/1, taken from specimen IZ-74–1 in the geological and paleontological collection of the Geopark-Trägerverein Braunschweiger Land–Ostfalen e.V). M. acutidens scales are of a transitional type between palaeoniscoid and lepidosteoid scales with a reduced dentine layer. b—bone, d—dentine, g—ganoine

Fig. 7

Mengius acutidens gen. et sp. nov. Paratype, specimen GG 23093. Left lateral view of the squamation at the dorsoposterior breaking edge. Arrows point to individual scales displaying the anterior margin. Specimen dusted with NH₄CL for photography

ZooBank LSID. zoobank.org:act:4327EBDD-6805-46EE-AD72-9430561DF7D4

1979Dapedus sp.—Herrig: Abb. 23.

1989 Lepidotes elvensis (Blainville, 1818)—Thies 1989b: 692–704, figs. 1–3

1991 Lepidotes elvensis (Blainville)—Ernst: Abb. 4

? 2007 Lepidotes elvensis—Ansorge: 41.

2015 Lepidotes elvensis (Blainville, 1818)—Ansorge and Obst: 235, Abb. 15d.

2016 Lepidotes elvensis (Blainville, 1818)—Ansorge and Grimmberger: 138, Abb. 12c.

? 2017 Lepidotes Agassiz 1832—Stumpf, Ansorge, Pfaff and Kriwet: 10.

2019 †Lepidotes sp. A, Specimen 2—Thies, Stevens and Stumpf: 2, fig. 2a–d

2019 †Lepidotes sp. B—Thies, Stevens and Stumpf: 2, 7, fig. 4c-d

Derivation of name. After Latin "acutus—pointed" and "dens—tooth".

Holotype. The specimen shown in Figs. 2 and 3. Repository and catalogue number: Geological collection of the Institute of Geography and Geology of the University of Greifswald, catalogue number GG 435.

Locality. The abandoned clay pit Klein-Lehmhagen ca 2 km NNE of the small town of Grimmen in the federal state of Mecklenburg-Western Pomerania in Northeast Germany. N54.131089, E13.059053. Data on the stratigraphy, lithology, and fauna of this locality were published by Ernst (1967, 1991), Ansorge (2007) and Ansorge et al. (2024).

Age. Early Jurassic, early Toarcian, falciferum Zone, exaratum Subzone.

Horizon. Grimmen Formation, Reinberg Member, exaratum concretions. The exact horizon is unknown.

Additional material. Paratype 1—The specimen shown in Figs. 4, 5 and 7. Repository and catalogue number: Geological collection of the Institute of Geography and Geology of the University of Greifswald, catalogue number GG 23093.

Locality. The abandoned clay pit of the former brickworks Schwinzer Ziegelei ca 3 km E of the small town of Dobbertin in the federal state of Mecklenburg-Western Pomerania in northeast Germany. N53.615833, E 12.111111. Data on the stratigraphy, lithology, and fauna of this locality were published by Ansorge and Obst (2007, 2015) and Thies et al. (2019).

Age. Early Jurassic, early Toarcian, falciferum Zone, ?exaratum Subzone.

Horizon. The specimen comes from a nodule within the Grimmen Formation (Reinberg Member) clay on top of the Posidonia Shales. The nodule agrees stratigraphically with the exaratum concretions in the section of the clay pit Klein-Lehmhagen.

Species diagnosis. As for genus by monotypy.

Stratigraphical range. So far known only from the Early Jurassic, early Toarcian, falciferum Zone, exaratum Subzone.

Description of the holotype. The specimen is an articulated, incomplete fish that lacks the posterior two thirds of the trunk and the dorsal, pelvic, anal, and caudal fins. The preserved portion of the fish measures 10.7 cm in length and 7.2 cm in maximum height. The fish is embedded in a marly nodule and acid prepared from the left side which is observable now (Fig. 2a). The anterior half of the skull was broken off, presumably before preparation. The parts of the nodule were subsequently glued together again, leaving a vertical fracture running from the skull roof down to a point behind the caudal end of the lower jaw. The cranial skeleton is preserved three-dimensionally and only slightly decomposed. Nearly all of the dermal cranial bones are preserved and still show their natural arrangement. The longitudinal distance from the anterior tip of the premaxilla to the posterior edge of the operculum measures 6.4 cm, the height of the skull from the posterior tip of the dentalosplenial to the dorsal midline at the medioposterior corner of the left parietal is 4.1 cm. The orbit has a maximum diameter of 1.7 cm. The distance tip of the snout to posterior edge of the operculum equals approximately four times the maximum diameter of the orbit. Only along the fracture small parts of damadged dermal ossifications were lost during preparation. Most postmortem destruction has affected the pectoral fin, which is preserved only fragmentarily.

Endochondral neurocranium—The endochondral neurocranium seems to be intact but is almost completely covered by dermal bones or unremoved matrix (in the orbit) and cannot be studied in the specimen by methods applied here. CAT scanning will probably yield insights into the structure of the endochondral neurocranium and is planned to be carried out in the future. Only a small part of what is probably the spenotic becomes visible in the gap between the left parietal and postparietal (Fig. 3, Spo) but does not yield any detailed information.

Dermal neurocranium

Rostral—The rostral is not preserved in the specimen.

Nasal (Fig. 3, Na)—The nasal is much higher than long and roughly crescent-shaped. The dorsorostral margin of the bone is strongly convex, the ventrocaudal border shows an irregular concave course. The external surface of the ossification is smooth. The bone is traversed by a canal for the anterior section of the supraorbital canal. The canal opens by two large pores near the dorsoanterior edge of the bone. On the outer face of the bone the canal appears as a slight elevation by which the course of the canal can be traced.

The nasal is a little displaced but seems to have had an original position relatively close to the midline so that both nasals were near to one another. The rostral and antorbital bones are lost in the specimen, therefore the exact shape of the narial opening remains obscure.

Parietal ("frontal", Fig. 3, Pa)—The parietal is the largest of the skull roofing bones. It borders the nasal rostrally and the postparietal caudally. Possibly the parietal also contacts the supratemporotabular caudally. Such a contact is, however, obscured by the fracture of the specimen running between the parietal, the postparietal and the supratemporotabular. Unfortunately, the posteriormost part of the parietal and the anterior parts of the postparietal and supratemporotabular broke off and became lost by the fracturing of the fossil. Laterally, the bone is sutured to the dermosphenotic caudally and to two supraorbitals rostrally. The lacrimal and the infraorbital 1 sunk with their dorsal ends inside the cranial cavity during decomposition of the specimen. Their suspicious height indicates, however, that in life they abutted against the rostrolateral edge of the parietal, as is seen in the skull reconstruction of Lepidotes elvensis by Wenz (1968: fig. 45). Medially, a slightly undulating suture indicates the connexion between left and right parietals.

The left parietal appears to be slightly longer than its right counterpart. It tapers only slightly anteriorly. In rostrocaudal extension it measures at least 3.1 cm, its mediolateral width reaches ca 1.1 cm at its posterior end. The length/width ratio is 2.8 (or a little more considering that the very caudal end of the bony plate is broken off).

The surface of the parietal is nearly plane and smooth lacking ornamentation. A series of pores along the lateral edge of the bone shows the course of the canal for the supraorbital sensory line (Fig. 3, pcsop). In the caudal half of the ossification a second, shorter series of pores opens on the surface (Fig. 3, pcsop). This series lies half way between the medial and lateral edges of the bone and runs parallel to them. These pores also belong to the supraorbital sensory line. At present it remains unclear whether there is only one canal for the supraorbital sensory line running parallel to the lateral edge of the bone or whether there is a rostro-mediad directed branch of the main canal deviating from it in the caudal half of the bone.

Postparietal ("parietal", Fig. 3, Ppa)—The postparietal is a plate-like ossification sutured to the parietal rostrally, to the supratemporotabular laterally and to the extrascapular caudally. Medially it borders its counterpart. The rostrolateral corner of the right postparietal is broken off and not preserved. The exact dimensions and shape of the bone remain uncertain. During decomposition of the specimen the left postparietal has slipped to some extent underneath its counterpart and the left extrascapular. The right postparietal is still covered by matrix laterally and caudally. The observable remains indicate a squarish shape.

The external face of the bone is smooth and devoid of any ornamentation. It does not become clear from the specimen whether the caudal end of the supraorbital canal enters the postparietal. The anterior pit line (Fig. 3, apl) is well developed and runs for a short distance nearly parallel to the lateral edge of the ossification. Caudally it meets the middle pit line (Fig. 3, mpl) in an acute angle. The middle pit line reaches the lateral border of the postparietal and continues onto the supratemporotabular. The posterior pitline (Fig. 3, ppl) is placed isolated and a little caudally and medially from the other two lines. It is curved and shorter than the anterior and middle pit lines.

Supratemporotabular ("dermopterotic", Fig. 3, Stta)—The left supratemporotabular is only partially preserved. Along the break line of the specimen large pieces of the rostral part of the bony plate are broken off and missing. The remaining fragment indicates that the supratemporotabular is an rostrocaudally elongated plate of bone which adjoins the parietal rostrally, the postparietal medially, the extrascapular caudally, the preoperculum caudolateroventrally, the dorsalmost suborbital lateroventrally, and the dermosphenotic rostrolateroventrally. As with the other skull roofing bones there is no ornament on the external surface of the bone.

Coming from the postparietal the middle pit line of the skull roof enters this bone on its caudal portion and almost traverses it mediolaterally. It terminates close to the lateroventral edge of the bone. Parallel to the lateroventral edge the supratemporotabular contains the canal for the otic section of the infraorbital line. A short section of the canal has become unroofed on the damaged rostral part of the ossification where superficial growth layers of bone were removed by breakage (Fig. 3, coio). On the caudoventral corner of the bone the postotic section of the cephalic lateral line opens by a large pore (Fig. 3, pcpocll).

Extrascapular (Fig. 3, Ex)—There is only a single pair of extrascapulars. These are tongue-shaped and mediolaterally slightly convex ossifications which meet with their taperings ends in the body midline. Rostrally they are in contact with the postparietals, laterocaudally with the posttemporals. Ventrally the extrascapular borders the operculum. Sparse tubercles grow on the external surface of the bony plate.

Two sensory canals are contained within the bone indicated by large pores which open on the external face of the extrascapular. One of these runs at the ventrolateral edge of the bone. It housed two different sections of the cephalic lateral line. The anterior portion of the canal between the rostrolateral corner of the bone and the branch-off of the supratemporal commissur enclosed the postotic section of the cephalic lateral line (Fig. 3, pcpocll). Posteriorly the canal for the cephalic lateral line continued to the caudal edge of the bone where it passes onto the neighbouring posttemporal (Fig. 3, pccll). The second canal is developed mediolaterally and housed the supratemporal commissure (Fig. 3, pcsc).

Antorbital—The antorbital is not preserved in the specimen.

Circumorbitals—The series of circumorbital bones comprises two supraorbitals and the dermosphenotic dorsally of the orbit and a lacrimal and the infraorbitals and suborbitals surrounding the orbit rostrally, ventrally and caudally. The eye is protected by slerotic bones.

Supraorbitals (Fig. 3, So)—The supraorbitals are oblong plates of bone with the rostral one being a little larger than the caudally situated plate. The rostral supraorbital forms the anterodorsal corner of the orbit. It is distorted longitudinally and deeply grooved externally. The mediodorsal edges of the supraorbitals are in contact with the frontal. The anterior suborbital has an expanded anteroventral portion and borders the infraorbital 2 rostroventrally. The posterior supraorbital is sutured to the dermosphenotic caudally. Their ventral edges are free. The posterior supraorbitals has ganoine tubercles sitting along the ventral margin of the external surface. In the anterior supraorbital the occurence of tubercles is constricted to the caudal portion of the ventral rim of the plate. The rostrodorsal edge of the anterior supraorbital shows a coarse denticulation.

Dermosphenotic (Fig. 3, Dsp)—The dermosphenotic is broken in pieces leaving only the rostral part of the bone undamadged. Its exact shape remains therefore uncertain. The intact rostral portion of the bone appears to have a thickened ventral edge. A single row of ganoine tubercles is strung along the the outer face of the ventral edge.

Three large pores along the dorsal edge of the bone indicate the presence of the infraorbital canal within the dermosphenotic (Fig. 3, pcio). The exact course of the canal remains, however, obscure. Pores on the neighbouring parietal suggest that there is an anastomosis between the infraorbital and supraorbital canals across the border between the dermosphenotic and the parietal, as in the living Amia (Grande and Bemis 1998: fig. 20).

Lacrimal ("infraorbital 1", Fig. 3, La)—Rostrally the circumorbital series terminates with the lacrimal. This is a plate-like, wedge-shaped ossification between the (missing) antorbital anteriorly and infraorbital 1 posteriorly. Ventrally it presumably contacted the dorsorostral rim of the maxilla which has become a little dislocated during decomposition of the specimen. The caudoventral corner of the bone touches the rostral extremity of the supramaxilla. Medially the lacrimal covers what appears to be the dermopalatine 1. It remains uncertain whether the lacrimal bordered with its rostromedial edge the parietal anterodorsally or if a gap separated the two bones.

The external surface of the lacrimal is smooth. Pores of the infraorbital canal open along the ventral rim of the bone (Fig. 3, pcio).

Infraorbitals (Fig. 3, Io)—The infraorbital series is completely preserved and includes nine plate-like elements arranged ventrally and caudally of the orbit. The numbering of the infraorbitals follows their subsequent appearance during ontogeny from anterior to posterior between the antorbital (lacrimal, respectively) and the dermosphenotic (López-Arbarello 2012). Except infraorbitals 6 and 7 they are all of roughly rectangular shape and of different sizes. The largest plate within the series is infraorbital 1 which does not take part in surrounding the orbit but is placed in front of it. It abuts the anteriormost portion of the parietal dosally, the lacrimal rostrally, and the infraorbital 2 caudally. Ventrally it touches the supramaxillary. Infraorbital 2 confines the orbit anteriorly. Dorsally, it should have contacted the anterior supraorbital. Further caudally the series continues with the infraorbitals 3 to 9 which delimitate the orbit ventrally and caudally with infraorbitals 7 to 9 forming the posterior rim of the orbit. Infraorbital 3 reaches the ectopterygoid ventrally whereas infraorbital 4 even covers it laterally to a small extend dorsocaudally. Infraorbitals 4 to 9 touch the neighbouring suborbitals below and behind. Ganoine tubercles are present on the external face of infraorbital 9 along the caudal and ventral margins of the plate. The other infraorbitals are smooth.

Pores belonging to the infraorbital canal open only on infraorbitals 1 to 6 (Fig. 3, pcio). The caudally placed infraorbitals 7 to 9 are free of pores. The course of the canal is indicated by a longitudinal swelling on the external surface of the plates near their rim which delimits the orbit.

Suborbitals (Fig. 3, Sob). The series of suborbitals ventrally and caudally to the infraorbitals comprises four plate-like, irregularly formed elements. The smallest of these ist the dorsalmost plate which abuts the supratemporotabular dorsally and the dermosphenotic and the dorsalmost infraorbital rostrally. Ventrad follows the largest plate in the series which is at least three times as large as the dorsal suborbital. It is in contact with the infraorbitals 7 to 9 in front. The suborbital placed near the ventrocaudal corner of the orbit touches the infraorbitals 6 and 7. Caudally the suborbitals cover the rostral portion of the preoperculum. The dorsal and the ventrocaudal suborbital carry some sparse tubercles on their external surface. The other two are smooth.

The suborbital in the ventrocaudal corner of the series shows a curvy groove on its external surface (Fig. 3, hpl). This groove represents the horizontal pit line of the cheek.

Below the middle of the orbit the ventral border of the circumorbital ring formed here by the ventral suborbital and the infraorbital 4 flexes abruptly dorsally. Further rostrally, at the level of infraorbital 2 it turned ventrally again.

Sclerotic bones (Fig. 3, Scb)—A further ossification of the orbit appears immediately at the posteroventral rim of the left anterior supraorbital. Here a curved, thin, ribbon-like lamina of bone sticks out of the unresolved matrix still filling the orbit. It continues for a short distance underneath the supraorbitals before it vanishes again in the matrix below the posterior supraorbital. This bone represents a sclerotic ossification. It remains uncertain whether the sclerotic ossifications comprise only a single bone or several individual bones or even a complete sclerotic ring.

Sclerotic bones have only rarely been reported to occur in ginglymodians, e.g. in Araripelepidotes (Thies 1996). They are presumably present in more ginglymodian taxa but were overseen so far because of insufficient preparation or investigation of specimens.

Parasphenoid—The parasphenoid or remains thereof cannot be identified in the specimen with certainty.

Vomer (Fig. 3, Vo)—Only the smooth dorsal face of the rostral part of the vomer is visible in the specimen. It remains obscure whether the vomer is an unpaired or co-ossified bone because the inside of the ethmoidal cavity is still filled with matrix mostly.

Visceral skeleton

Mandibular arch

Endochondral elements

Quadrate (Fig. 3, Qu)—The quadrate is an individual ossification not beeing co-ossified with the metapterygoid, entopterygoid and ectopterygoid. In the specimen a massive condyle for the articulation of the articular in the lower jaw becomes visible but most of the quadrate is still embedded in matrix. The position of the condyle indicates that the lower jaw articulation is situated at the level around the centre of the orbit. This becomes also obvious from the paratype (Fig. 5, Qu).

Articular (Fig. 3, Art)—This bone is covered mostly by the dermal ossifications of the lower jaw and and only its caudalmost part including the articular facet for the quadrate condyle emerges underneath the angular.

Dermal elements

Premaxilla (Fig. 3, Pmx)—The paired premaxilla forms the tip of the snout. It comprises a mediolaterally elongated tooth-bearing ventral part and a large nasal process arising from the medial end of the ventral part dorsocaudad into the nasal pit, which is still filled with matrix.

The ventral part is a dorsoventrally flattened plate of bone with a convex anterior border. It tapers laterally and originally bore a row of seven high, styliform and pointed teeth on its ventral surface, but only five of which are preserved. Together with the dentary teeth in the lower jaw the premaxillary teeth represent the largest teeth in the dentition of GG 435 (Fig. 2b).

The nasal process is a thin lamina of bone sutured to the rostrocaudally widened medial end of the tooth-bearing part of the premaxilla. It is approximately one and a half times as high as the ventral part of the premaxilla is wide, and lines the nasal pit. Two foramina are present on the nasal process. One of these is small and opens at the ventral base of the process immediately above the suture line between the ventral portion of the premaxilla and the nasal process (Fig. 3, r. pal. VII). Similar foramina are present on the premaxillae of, e. g. Macrosemimimus lennieri (Sauvage, 1893) from the Upper Jurassic of France and Callipurbeckia minor (Agassiz, 1833) from the Upper Jurassic/Lower Cretaceous of England and France and are considered as an opening for the terminal branch of the palatine nerve (Wenz 1968: 85, fig. 39; Patterson 1975: fig. 135).

A second foramen opens further dorsally. It is much larger than the ventral one and of slot-like outline. It occupies nearly the entire dorsal half of the nasal process (Fig. 3, N I). In size, shape and position it is comparable to a foramen occurring in the premaxillary nasal processes of, e. g. Macrosemimimus lennieri (Sauvage, 1893) and Lepidotes elvensis (Blainville, 1818) from the Upper and Lower Jurassic, respectively, of France and Pliodetes nigeriensis Wenz, 1999 from the Lower Cretaceous of Nigeria (Wenz 1968: 85, 97, figs. 39, 44; 1999: 113, fig. 4). In the living Amia this foramen serves for the passage of the olfactory nerve (Grande 1998: 81, fig. 42 C, D). The homology of the premaxillary nasal process in neopterygians was dicussed by López-Arbarello (2012).

Maxilla (Fig. 3, Mx)—As seen in lateral view the maxilla is a wedge-shaped bony plate. It reaches its maximum height in its caudalmost part and tapers anteriorly where it presumably abutted the lacrimal with its dorsorostral edge. Its anteriormost extremity is bent medially under the lateral tip of the premaxilla to form a process. By this process the maxilla was anchored to and articulated with the ethmoidal neurocranium which is still covered by matrix in the specimen. The caudal end of the maxilla resting on the dorsocaudal part of the dentalosplenial rotated around this anterior process when the jaws were opened or shut.

The left maxilla originally bore a series of 17 delicate teeth along its ventral edge of which only 12 are preserved. All the teeth are styliform, slender, unicuspide and pointed. Their tips are curved medially more or less. They reach only two thirds or less of the height of the premaxillary teeth.

The outer surface of the maxilla is smooth. Some pores are present on this surface with one of them being remarkedly large and slot-like. These pores were considered by Thies (1989a, b) as traces of a maxillary pit line (Fig. 3, pmpl).

Supramaxilla (Fig. 3, Smx)—There is only one supramaxilla present in GG 435. The bone is rostrocaudally elongated and of plate-like appearance. The plate is smooth externally and of slightly boomerang-shaped outline with a convex dorsal and a concave ventral rim. Ventrally the ossification was sutured to the middle and posterior part of the dorsal edge of the maxilla. Medially it rests on the ventral part of the ectopterygoid.

Ectopterygoid (Fig. 3, Ectp)—Only the dorsal portion of the ectopterygoid is visible in the gap between the infraorbitals 1 to 4 and the ventral suborbital dorsally and the supramaxilla, the surangular and the angular ventrally. It displays a smooth lateral surface. According to the visible part of the bone the ectopterygoid is a rostrocaudally elongated plate or strip of bone which contacts the dermopalatine rostrally and the quadrate caudally. This agrees with the condition in Lepidotes elvensis (Blainville, 1818) from the Lower Jurassic of France (Wenz 1968: fig. 47). In L. elvensis the ectopterygoid possesses teeth rostrally. It remains obscure whether this is also the case in GG 435 because the rostroventral part of the ectopterygoid is covered by the supramaxilla.

Dermopalatine (Fig. 3, Dpal)—The dermopalatine is still mostly enclosed by matrix. Only the anterior part with one or two conical teeth is free but can hardly be observed because of overlying portions of the lacrimal, premaxilla and maxilla.

Dentalosplenial (Fig. 3, Dspl)—The lateral cover of the lower jaw is formed by several ossifactions. One of these is the dentalosplenial which extends rostrocaudally from the tip of the mandible to almost over the entire length of the mandible. It reaches its maximum height in the caudal part of the bone where it takes part in the formation of the coronoid process, together with the surangular and the angular. Specimen GG 435 presents the left dentalosplenial. The right dentalosplenial is still buried in the matrix. Only the symphyseal tip of it emerges (Fig. 3, Dspl(r)).

Below and behind the coronoid process the dentalosplenial forms a slender, pointed and caudally directed process which ends a little in front of the articulation of the lower jaw with the quadrate. Caudally and above the caudal process the dentalosplenial is joined to the surangular and angular.

The dentalosplenial bears high, slender and unicuspid teeth on the anterior dorsal edge (Fig. 2b). Their pointed apices are slightly curved to the interior of the oral cavity. In shape and size the dentary teeth are comparable to those of the premaxilla. Nine dentary teeth aligned in a single row are preserved in GG 435. Gaps in the row indicate to five missing teeth, which were either shed before death or broke off during preparation of the specimen.

The lateral surface of the bone is smooth. A double row of pores is developed longitudinally on the anterior part of the bone beginning at the tip of the mandible. The dorsal row (Fig. 3, poc) is shorter, has smaller pores and belongs to the oral canal. The ventral row (Fig. 3, pmdc) is longer and extends to the ventrorostral corner of the angular where it continues. Its pores are larger and indicate the course of the mandibular canal.

The formation and function of the oral canal was previously studied by Wenz (1968), Thies (1989a, b), Cavin and Suteethorn (2006) and Leuzinger et al. (2019). Cavin and Suteethorn (2006) found the oral canal to be present in different patterns of formation in a series of fossil ginglymodian taxa. They considered its presence as a derived character. Leuzinger et al. (2019) studied this canal in an isolated lower jaw of the lepisosteiform neopterygian Scheenstia sp. from the Upper Jurassic of Switzerland using μ-CT scanning technique. They confirmed that the oral canal is not connected to the mandibular sensory canal. According to Leuzinger et al. (2019) the oral canal possibly served for the passage of rami of the trigeminal–facial nerve innervating (hypothetical) dermal sensory structures such as thickened lips or barbels.

Angular (Fig. 3, Ang)—The angular is a slightly vaulted plate of bone of irregular outline with concave rostral and convex caudal borders. Rostrally it is sutured to the dentalosplenial and dorsally to the surangular. The lateral surface of the bone is smooth except of three shallow, flaky, oval-shaped ganoine coverings near the ventral border of the bone.

Coming from the dentalosplenial the mandibular canal enters the angular at its rostroventral corner. The canal continues caudally parallel to the ventral edge of the angular as indicated by a series of large pores opening along this edge (Fig. 3, pmdc). It leaves again at the caudal edge of the bone a little dorsally of its caudoventral corner. Additionally, the ventrocaudal portion of the angular bears the mandibular pit line (Fig. 3, mdpl). The line takes a boomerang-shaped course from the ventral to the caudal border of the ossification.

Surangular (Fig. 3, Sang)—The surangular forms the dorsal tip of the coronoid process. Ventrally it is sutured in a straight line to the angular. The rostral and dorsocaudal edges of the ossification are free. The visible part of the bone has the outline of an irregular triangle. The lateral face of the angular is smooth.

Coronoids (Fig. 3, Co1, Co2)—Two coronoids are visible in GG 435. Both appear to belong to the left mandible where they originally lined the medial surface of the dentalosplenial. During decomposition of the specimen they have become displaced from the lower jaw.

The coronoids are thin plates of bone which bear styliform, unicuspid teeth on their medial surface. The exact shape and the dimensions of the plates remain obscure because they are still partly embedded in matrix. Coronid 1 seems to be a little smaller than coronid 2 which is more elongated and caudally covered by the dentalosplenial and the maxilla. The teeth of coronid 1 are arranged in two groups. On the (supposed) dorsal edge of the ossification is a series of 10 teeth which correspond in morphology to the premaxillary or dentary teeth, but reach only half of their size. Placed more centrally on the plate stand six more teeth which are a little taller and much stouter than the marginal teeth. On coronid 2 only the dorsal marginal row of teeth is observable. The series comprises at least 12 teeth which agree in morphology and size with the marginal teeth on coronid 1 (Fig. 2b).

The prearticular cannot be observed in GG 435.

The dentition is non-tritoral. The teeth on premaxilla and dentalosplenial are of piercing-type suited (amongs others) to grasp and hold a moderately mobile prey.

Hyoid arch

Endochondral elements

The endochondral elements of the hyoid arch including the hyomandibula are almost entirely concealed by the dermal bones of the cheek and hyoid arch

Symplectic (Fig. 3, Sym)—The symplectic connects the quadrate with the hyomandibula and in this way gives additional support to the lower jaw articulation. A piece of an endochondral rod-like bone comes out of unresolved matrix from underneath the quadrate condyle. In shape, position and orientation this bone roughly agrees with a symplectic. A symplectic is present in specimens of "Lepidotes sp." from the Upper Jurassic of England (Patterson 1973: fig. 26 B, C). In the specimens concerned the symplectic is also of rod-like appearance and braces the lower jaw articulation by couching the posteroventral edge of the quadrate. The specimens named "Lepidotes sp." by Patterson (1973) probably represent a mixture of specimens belonging to Macrosemimimus and Callipurbeckia (López-Arbarello 2012; Schröder et al. 2012).

Questionable ceratohyal or hypohyal (Fig. 3, ?Ch/?Hh)—Otherwise, only one further element of the left endochondral hyoid arch becomes partially visible in the gap between the caudal edge of the left mandible and the rostral tips of the preoperculum, interoperculum and the branchiostegal rays. Because of its position this part of bone could represent a ceratohyal or hypohyal.

Dermal elements

Operculum (Fig. 3, Op)—The operculum is roughly semi-circular in outline. The dorsal, caudal and ventral borders are convex and merge smoothly into one another. The rostral edge is straight or slightly concave. Ventrally the operculum overlapps the suboperculum, rostrally it abuts the preoperculum. Dorsocaudally the operculum rests against the supracleithrum, but caudally it covers the cleithrum completely, because the entire skull was shifted caudally during fossilisation. The lateral surface of the operculum is smooth.

Suboperculum (Fig. 3, Sop)—The suboperculum forms the ventrocaudal corner of the opercular apparatus and is of roughly trapezoidal outline. The bone shows a convex caudoventral margin. It borders the operculum with a concave and the interoperculum with an almost straight suture line. The rostrodorsal corner of the suboperculum is formed into a short, dorsally directed process which extends alongside the rostral edge of the operculum for a ninth of its length. As in the operculum the lateral face is smooth.

Interoperculum (Fig. 3, Iop)—The interoperculum is tongue-shape in lateral aspect. Its maximum length equals nearly two times its maximum height at the caudal border. Its rostral tip is directed to the caudoventral corner of the mandible. The dorsal part of the lateral surface articulates with the medial surface of the preoperculum. As in the operculum and suboperculum the lateral surface of the interoperculum is smooth.

Branchiostegal rays (Fig. 3, Rb)—In the figs. 1 and 2 in Thies (1989a, b) GG 435 shows remains of the left branchiostegal ray apparatus comprising six elements. Meanwhile the proximal part of the ventralmost ray is broken off and lost. The distal tip is still preserved. All of the rays are rostrocaudally elongated plates of bone of sabre-like shape. They increase in length and distal height ventrally to dorsally. The rays articulate proximally with the presumed ceratohyal or hypohyal. The two dorsal rays are largely covered by the interoperculum and suboperculum. The lateral surface of the rays is smooth. Together they form a fan-shaped structure. A narrow strip-like bone laying alongside the ventral edge of the dentalosplenial represents an isolated element from the brachiostegal series of the right side of the skull (Fig. 3, Rb(r)).

No gular plate or remains thereof are preserved or visible in the specimen.

Preoperculum (Fig. 3, Pop)—The preoperculum is a narrow and curved ossification of boomerang-shaped appearance. It extends from the supratemporotabular dorsally to the rostral tip of the interoperculum ventrally without developing clearly defined dorsal and ventral arms. The rostral part of the bone articulates with the caudal and ventrocaudal margins of the suborbitals in front. The caudal edge borders the operculum, suboperculum and interoperculum behind.

The preoperculum houses the canal for the preopercular sensory line. The preopercular canal opens by a series of large, slit-like pores only on the ventral arm of the bone (Fig. 3, ppc). The pores are lined up along a longitudinal ridge. Caudally and ventrocaudally of this ridge the external surface of the preoperculum is wavy, otherwise it is smooth.

Branchial apparatus

The branchial apparatus or parts thereof are not visible in the specimen

Axial skeleton

Remains of the axial skeleton are concealed underneath the scale armour in GG 435 and cannot be observed.

Shoulder girdle and pectoral fins

Endochondral shoulder girdle

Endochrondral elements of the shoulder girdle cannot be identified on the specimen

Dermal shoulder girdle

Posttemporal (Fig. 3, Pt)—The posttemporal connects the dermal shoulder girdle with the skull roof. The exact dimensions and shape of the bone remain, however, obscure because its caudal and mediocaudal borders are still covered by matrix. As far as can be judged from the observable remains the posttemporal is of roughly triangular outline and borders the ventral part of the extrascapular rostrally. Dorsally and caudally it probably contacts the scale armour. The ventrolateral margin of the posttemporal is free rostrally, its relationship to the operculum is obscured because of slight dislocation of both ossification during fossilisation. Caudally the ventrolateral edge of the posttemporal is shaped into a sigmoidal facet where the supracleithrum articulates with the posttemporal by a wide dorsal process.

The posttemporal is traversed by the canal for the cephalic lateral sensory line. The exact course of the canal remains obscure, it opens, however, on the external surface of the bone by two large pores near its ventrocaudal corner (Fig. 3, pccll). There are no ganoine tubercles present on the outer surface.

Supracleithrum (Fig. 3, Scl)—Like the posttemporal the supracleithrum is of roughly triangular shape. The dorsal edge of the bone has a broad longitudinal process which articulates with the opposed facet on the ventral border of the posttemporal. The caudal border of the bone is nearly straight. The rostroventral border is overlapped a little by the operculum which rests against the supracleithrum.

The supracleithrum houses the posteriormost portion of the cephalic lateral sensory line. The canal for the sensory line opens by three large pores near the dorsal edge of the bone (Fig. 3, pccll). There is no ornamentation on the external face of the supracleithrum.

Cleithrum (Fig. 3, ?Cl)—The cleithrum is nearly completely concealed by the opercular apparatus which was dislocated a little caudally during fossilisation. Only a small strip of bone emerging from underneath the ventrocaudal edges of the suboperculum and one of the branchiostegal rays probably belongs to the cleithrum. Not more can be said on this element of the dermal shoulder girdle.

Postcleithra (Fig. 3, Pcl)—There are two postcleithra present behind the cleithrum. A third one may have filled the gap between the ventral postcleithrum, the possible part of the cleithrum and the pectoral lepidotrichia but was lost during decomposition or preparation of the specimen. In having their caudal and ventral edges slightly curved the preserved two postcleithra resemble the caudally neighbouring scales in shape but are of much larger size. In height the dorsal postcleithrum equals three and a half times the scales abutting it posteriorly. The lateral surface of the postcleithra is smooth.

Pectoral fin

The specimen displays the left pectoral fin of which only the proximal part is preserved (Fig. 2a). Its exact shape and size therefore remains obscure. The preserved part includes at least 10 unsegmented basal pieces of lepidotrichs. The leading edge of the pectoral fin is protected by fulcra.

The fin is placed very low on the flank near the ventral midline of the body behind the cleithrum. Endochondral elements of the pectoral fin skeleton are not preserved in GG 435.

Pelvic girdle and fins

The specimen is broken transversely near the middle of the trunk with the posterior part of the body not being preserved. Nothing can be said therefore on the structure and position of the pelvic girdle and fin.

Unpaired fins

Remains of the unpaired fins are not preserved in the specimen.

Scales

All scales display their lateral surface, and in nearly all scales only the free field is observable. In shape they vary according to their position on the trunk. Scales in the centre behind the postcleithra are rhombohedral in outline with slightly curved caudal and ventral margins (Fig. 2a). In these scales the height of the free field exceeds its length with the height/length ratio taking a value of approximately 5:3 (mm). To the dorsal and ventral midline, the scales decrease in size and become gradually rhomboidal in outline.

In most scales the caudal margin is damaged. In some intact scales the caudal margin is coarsely serrated. In a scale immediately posterior to the ventral postcleithrum the serration includes at least five widely spaced denticles (Figs. 2a, 3a). Further caudally and dorsally serration becomes gradually reduced. This concerns the number of denticles as well as restriction of their occurrence on the ventral part of the caudal scale margin. Finally, only the ventrocaudal corner of the scales extends as a little pointed tip.

The scales in GG 435 have a vertical peg-and-socket articulation. In at least some species of "Lepidotes" occurs an additional longitudinal articulation by anterior dorsal and ventral scale processes which are overlapped by adjacent scales (e.g. Scheenstia mantelli (Agassiz, 1833), Callipurbeckia minor (Agassiz, 1833); López-Arbarello 2012). An anterior dorsal process which is even larger than the dorsal peg is present in the scales of the GG 435. The development of an anterior ventral process in the scales remains, however, obscure in the specimen because none of the scales of GG 435 displays the anterior edge or medial side sufficiently in order to confirm the presence of such a process.

Thin sections of the rhomboid scales of specimen GG 435 (Fig. 6) display a layered osteocytic bony base in part overlain by a reduced dentine forming ridges. The scales show well-developed canals of Williamson. The whole scale is covered by layered ganoine. The scales of Mengius acutidens gen. et sp. nov. are therefore similar to those of Lepidotes and transitional between the palaeoniscoid and lepidosteoid type (Schultze 1966).

Specimen GG 23093—additional information and variation. Additional information on the skeletal morphology and anatomy of Mengius acutidens gen. et sp. nov. and evidence of individual variation is given by specimen GG 23093 (Figs. 4, 5, 7). This concerns the following aspects:

Postparietal ("parietal", Fig. 5, Ppa)—In specimen GG 435 the postparietal is damaged and does not yield any hints for the posterior portion of the supraorbital sensory canal being included in it. Possibly this section of the supraorbital line is replaced in the holotype by the anterior pit line of the skull roof (Fig. 3, apl). Specimen GG 23093 has a better preserved left postparietal which shows a short series of pores along its anterior lateral margin (Fig. 5, pcsop) indicating that in this specimen a short canal for the posterior section of the supraorbital sensory line is formed within the postparietal parallel and near its lateral edge. It remains unclear whether there is an anterior pit line on the external surface of the postparietal of GG 23093 which would represent a continuation of the supraorbital sensory line posteriorly.

Dermosphenotic (Fig. 5, Dsp)—The dermosphenotic is well preserved in GG 23093 on both sides of the cephalic skeleton. The bone is of irregular shape with slightly curved dorsal and rostral borders. Both dermosphentics show a slot-like groove parallel to the dorsal edge of the bone which in the left dermosphenotic is still filled with sediment. On the right dermosphenotic several foramina open into this groove. As in the holotype these might indicate an anastomosis within the dermosphenotic of the supraorbital and infraorbital canals, as in the living Amia (Grande and Bemis 1998: fig. 20). The posterior section of the supraorbital canal runs near the ventral border of the parietal and postparietal as shown by its pores (Fig. 5, pcsop). The infraorbital canal leaves the dermosphenotic caudally and continues into the supratemporotabular which bears a series of small pores on its outer face belonging to the postotic section of the infraorbital canal (Fig. 5, pcoio). A single large pore near the caudoventral corner of the dermosphenotic (Fig. 5, pcio) marks the transition of the infraorbital canal into the series of circumorbitals starting dorsally with infraorbital 9 which is also pierced by a pore of the infraorbital canal laterally.

The lateral surface of the dermosphenotic is irregularly corrugated. A prominent structure is a deep and narrow vertical groove which cuts into the caudoventral edge of the bone but widens and flattens out rostrodorsally. This morphology is equally present on the left and right dermosphenotic.

Infraorbitals (Fig. 5, Io)—The series of infraorbitals is completely preserved on the left side of the specimen (Figs. 4a, 5). As in the holotype it includes nine elements which are of similar shape and size as in GG 435 (Fig. 3, Io). On the right side only six infraorbitals are preserved anteriorly and ventrally of the orbit (Fig. 4b). The infraorbitals posterior of the orbit are missing but the distance between the caudalmost preserved infraorbital (infraorbital 6) and the dermosphotic agrees with the space taken by the infraorbitals 7 to 9 on the left side of the specimen. In the holotype the same area is equally filled by the infraorbitals 7 to 9 (Fig. 3, Io7-Io9). It may therefore be assumed that on the right side of the skull of GG 23093 the series of infraorbitals was also completed by three further infraorbitals posterior to the orbit.

Suborbitals (Fig. 5, Sob)—Only one element of the suborbital series is preserved in GG 23093. This is the largest element of the series placed between the infraorbitals 6 to 9 and the dorsal arm of the preoperculum. No other suborbitals are preserved in GG 23093 on both sides of the skull. It remains uncertain therefore how many suborbitals were present in the specimen originally. The suborbital preserved in GG 23093 is higher and larger than the corresponding ossification in GG 435 (Fig. 3, Sob). In height it reaches the dorsal end of the preoperculum whereas in GG 345 there are two suborbital plates posterior to the orbit, a small one being apposed to the dorsal end of the preoperculum followed ventrally by a large one which extends along most of the dorsal arm of the preoperculum. It seems therefore that the posterior suborbital in GG 23093 has resulted from fusion of two suborbital plates which occur in the holotype. Also, the dorsocaudal corner of the suborbital in GG 23093 is formed into a spine-like posterior process on both sides of the skull.

Metapterygoid (Fig. 5, Mtp)—The metapterygoid becomes visible on both sides of the specimen and can be studied best on the right side of the cephalic skeleton (Fig. 4b). On this side all of the covering cheek bones are dislocated or missing so that the metapterygoid lies mostly free. In lateral aspect it is a smooth, concave plate of bone of roughly semicircular shape. On the left side it is still partly covered dorsally by the infraorbitals 5 and 6 and by the large posterior suborbital plate (Fig. 5, Mtp). The ventral suborbitals are also missing here. Caudally the edge of the metapterygoid slides under the preoperculum on the left side. Ventrally the metapterygoid itself covers a bony plate which might represent an entopterygoid (Fig. 5, ?Entp). In the Upper Jurassic Macrosemimimus lennieri (Sauvage, 1893) and “Lepidotes sp.” from France and England the metapterygoid has a very similar outline (Wenz 1968: figs. 41, 42; Patterson 1973: fig. 26B).

Questionable entopterygoid (Fig. 5, ?Entp)—A part of a bony plate which is wegded in between the metapterygoid dorsally and the quadrate ventrally is difficult to identify. Its position ventral to the metapterygoid suggests that it might be a part of the entopterygoid. A similar position is taken by the entopterygoid in the Upper Jurassic Macrosemimimus lennieri (Sauvage, 1893) from France (Wenz 1968: fig. 41).

Quadrate (Fig. 5, Qu)—In GG 23093 the entire quadrate is preserved and observable. The specimen confirms that the quadrate is not fused to the palate ossifications. The dorsal edge of the quadrate is free and nearly semicircular in lateral aspect. The lateral surface of the bone is deeply concave with the deepest point just above and a little behind the condyle for the lower jaw articulation.

Quadratojugal (Fig. 5, Quj)—An independent splint-like bone with a flattened and slightly widened rostral end extends along the ventral edge of the quadrate. Posteriorly it projects considerably beyond the quadrate and connects the quadrate with the preoperculum. Its flattened and widened anterior portion covers the quadrate condyle laterally and supports the lower jaw articulation. In position, shape and dimensions it strongly resembles the independent quadratojugal Patterson (1973: 292–293, fig. 26B) identified in “Lepidotes sp.” from the Upper Jurassic of England. In Macrosemimimus lennieri (Sauvage, 1893) the quadratojugal appears to be fused to the quadrate anteriorly and was interpreted as a posterior process of the quadrate (Wenz 1968: 91, fig. 41). In Macrosemimimus fegerti Schröder, López-Arbarello and Ebert, 2012 from the Upper Jurassic of Germany the quadratojugal is an individual ossification not fused to the quadrate anteriorly (Schröder, López-Arbarello and Ebert 2012).

Dentalosplenial (Fig. 5, Dspl)—Even though the angular has slightly shifted posteriorly during decomposition of the specimen the lower jaw agrees in formation and structure with the lower jaw of the holotype (Fig. 3, Dspl). The dentition of the dentalosplenial is only badly preserved with most of the teeth being destroyed. One complete tooth is, however, still present. It is tall, slender and pointed, and agrees in morphology well with the dentalosplenial teeth of the holotype.

Opercular apparatus (Fig. 5; Op, Sop, Iop)—In dimensions and shape the elements of the opercular apparatus agree with those of the holotype (Fig. 3; Op, Sop, Iop).

Branchiostegal rays (Fig. 5, Rb)—In contrast to the holotype there are more branchiostegal rays observable in GG 23093. Whereas there are only six rays present in GG 435 (Fig. 3, Rb(l)) specimen GG 23093 shows nine of them on the left side of the skull. In shape, size and morphological variation in ventrodorsal direction these nine rays agree with the rays preserved in the holotype and form a complete branchiostegal apparatus.

Postcleithra (Fig. 5, Pcl)—Whereas the holotype has only two dorsal postcleithra preserved (Fig. 3, Pcl) GG 23093 shows that the complete series of postcleithra comprises four elements. The two dorsal postcleithra are similar in size and outline in GG 435 and GG 23093. In GG 23093 these are slightly better preserved and the ventral one shows remains of a posterior serration. More ventrally and rostrally two more postcleithra are observable in GG 23093. These are much smaller than the dorsal ones, of triangular and elongated shape, repectively, but have also smooth lateral surfaces.

Scales (Figs. 4a, 7)—GG 23093 shows only three scales of which the anterior border is observable (Fig. 7). These are placed at or near the dorsocaudal fracture edge of the fragmentary specimen. The scales in question have only an anterodorsal process but lack an anteroventral process in the sense of López-Arbarello (2012).

Specimen GG 23093 allows to identify the number of scales in individual vertical scale rows (Fig. 4a). 21 scales are present in the first complete dorsoventral row of scales immediately behind the pectoral girdle. One scale near the ventral midline seems to be missing, so the entire amount should have been 22 scales originally. The second vertical row comprises 24 scales. Scale rows further caudally are more or less damaged dorsally and ventrally, and exact number of scales cannot be determined therefore.

Discussion. It becomes clear from the descriptions and figures above that the two fish specimens from the lower Toarcian of Mecklenburg and Western Pomerania belong to the same genus and species. With regard to the structure of the cephalic skeleton they resemble the contemporaneous "Lepidotus" buelowianus as described and figured by Jaekel (1929). Unfortunately, the holotype and only specimen is lost. A comparison can therefore be made only using Jaekel’s (1929) description and figure of the specimen. It should have measured 12 cm in longitudinal dimension, but a scale bar is lacking in Jaekel’s (1929) fig. 1 (here reproduced in Fig. 1). The teeth are described by Jaekel (1929: 17) as styliform with the apex being a little bulbous ("(Der Mund) …läßt in dem mittleren Teil des Dentale des Unterkiefers 7 Stiftzähne mit der etwas kolbenförmig verdickten Spitze erkennen."). Further resemblance exists in the number of teeth on the dental, premaxilla and maxilla, even though these numbers are slightly smaller in GG 435. On the other hand Jaekel’s (1929) fig. 1 reveals considerable differences in the cephalic skeleton between "L." buelowianus and our specimens:

• Ratio of "maximum diameter of the orbit" to "distance from tip of the snout to posterior edge of the operculum": This ratio equals approximately 1:6 in "L." buelowianus whereas it is 1:4 in GG 435 and 1:3,8 in GG 23093.

• Ventral border of the operculum: This border is slightly concave in "L." buelowianus whereas it is convex in GG 435 and GG 23093.

• Maxilla: The maxilla is proportionally longer and less high in "L." buelowianus.

• Dorsal suborbital behind the orbit: This element of the suborbital series is nearly as high as long in "L." buelowianus but much higher than long in our specimens, in particular in GG 23093.

• Size of dental teeth: Dental and maxillary teeth appear to be of equal size in "L." buelowianus whereas in GG 435 the dental teeth are nearly twice as high.

• Number of infraorbitals: There are eight infraorbital bones in "L." buelowianus but nine in our specimens.

We consider these differences as substantially enough to conclude that Jaekel’s specimen of "L." buelowianus and our specimens represent different species.

Our specimens bear as well resemblance to Lepidotes gigas Agassiz, 1832 und L. elvensis (Blainville, 1818) from the Toarcian of Northwest Europe. Also, many Jurassic neopterygians previously included in Lepidotes Agassiz, 1832 are attributed today to a number of different genera comprising amongs others Lepidotes Agassiz, 1832, Scheenstia López-Arbarello and Sferco, 2011, Callipurbeckia López-Arbarello, 2012, Macrosemimimus Schroeder, López-Arbarello and Ebert, 2012, Isanichthys Cavin and Suteethorn, 2006, and Camerichthys Bermúdez-Rochas and Poyato-Ariza, 2015. In order to clarify the systematic identity of the two specimens from Mecklenburg and Western Pomerania they will be compared to representatives of these six genera.

Lepidotes Agassiz, 1832 (species included L. gigas Agassiz, 1832, L. elvensis (Blainville, 1818), L. semiserratus Agassiz, 1837).

With the species presently considered to belong to Lepidotes all awaiting revision, the genus is still insufficiently diagnosed. The diagnostic characters tentatively listed by López-Arbarello (2012) comprise amongst others the characters (1) first anterior infraorbital bone deeper than more posterior anterior infraorbitals, (2) approximately squared-shaped infraorbital bones forming the posterior border of the orbit, (3) peg-and-socket articulation reduced or absent, (4) middle pit line contained in a groove excavated in supratemporotabular (= "dermopteroticum") and postparietal (= "parietal"), (5) the presence of a single pair of extrascapular bones, (6) numerous suborbital bones of variable size and shape, arranged in a series, which extends ventral to the orbit covering the quadrate laterally, and (7) thick ganoid scales with strongly developed longitudinal articulation through large dorsal and ventral anterior processes.

Of these only characters (1), (4) and (5) agree with our specimens. With regard to the remaining characters they disagree. The posterior infraorbital bones are non-squared-shaped, the peg-and-socket articulation is well developed, there are only four (or in the paratype possibly only three) suborbital bones, and the scales show only a dorsal anterior process. Further diagnostic characters listed by López-Arbarello (2012) refer to body measurement which cannot be applied to our specimens because of their incomplete preservation.

Important are, however, two characters which were not considered by López-Arbarello (2012): type of dentition and formation of ventral edge of operculum.

• Dentition: In L. gigas, L. elvensis and L. semiserratus the dentition is tritoral (López-Arbarello 2012: 27, fig. 20), but in the two specimens from Northeast Germany it is non-tritoral (Fig. 2). Figure 4 in Thies et al. (2019) also illustrates strikingly the differences in dental morphology between Lepidotes and our specimens.

• Ventral edge of the operculum: In L. gigas, L. elvensis and L. semiserratus this edge is straight or nearly so (López-Arbarello 2012: fig. 20, Woodward 1897: pl. 47, fig. 1). By contrast,  it is convex in the specimens from the Toarcian of Northeast Germany (Figs. 2, 4).

From the differences in cranial skeletal morphology it becomes evident that the two specimens from the Toarcian of Mecklenburg and Western Pomerania are neither conspecific with one of the species of Lepidotes nor that they belong into this genus at all.

Scheenstia López-Arbarello and Sferco, 2011 (species included S. maximus (Wagner, 1863), S. decoratus (Wagner, 1863), S. laevis (Agassiz, 1837), S. mantelli (Agassiz, 1833), S. zappi López-Arbarello and Sferco, 2011, S. degenhardti (Branco, 1885), S. bernissartensis (Traquair, 1911), S. hauchecornei (Branco, 1887)).

Scheenstia was diagnosed in detail by López-Arbarello (2012). Diagnostic characters include amongst others (1) three or more pairs of extrascapular bones, (2) dentition extremely tritoral, (3) orbital sensory canal present, (4) infraorbitals at the posterior border of the orbit longer than deep, and (5) thick ganoid scales … with very well-developed longitudinal articulation through large dorsal and ventral anterior processes.

None of these characters are present in our specimens. They can therefore not be assigned to Scheenstia or one of its currently known species.

It should be noted that in the cladistic analyses of López-Arbarello and Sferco (2018, 2020) the genus Scheenstia appears as a non-monophyletic (paraphyletic) taxon. The reason for this inconsistency remains obscure because the authors do not discuss the relationships of the taxa within the Lepidotidae in detail.

Callipurbeckia López-Arbarello, 2012 (species included C. minor (Agassiz, 1833), C. notopterus (Agassiz, 1837), C. tendaguruensis (Arratia and Schultze, 1999)).

In her diagnosis of Callipurbeckia López-Arbarello (2012) combined morphometric characters, such as body dimensions and relative fin positions together with osteological features. Body dimensions and fin positions cannot be determined in our specimens because of their fragmentary nature. Osteological characters comprise (1) skull bones ornamented with tubercles, (2) single pair of extrascapular bones, (3) two suborbital bones, a small oval dorsal suborbital and a much larger ventral suborbital filling most of the area between the infraorbitals and preoperculum, (4) maxilla deep, forming a more or less circular plate, (5) dentition moderately tritoral, (6) conspicuous dorsal ridge of scales, and (7) ganoid scales with well-developed vertical and longitudinal articulation with large dorsal peg and large dorsal and ventral anterior processes.

Of these only characters (1) and (2) are present in the two specimens from the Toarcian of Northeast Germany. Characters (3) to (7) are formed in a different way. Apparently, the two specimens do belong to Callipurbeckia or its species neither.

Macrosemimimus Schroeder, López-Arbarello and Ebert, 2012 (species included M. fegerti Schroeder, López-Arbarello and Ebert, 2012, M. lennieri (Sauvage, 1893)).

The genus is diagnosed by a combination of morphometric and osteological characters. Again, the morphometric characters comprising body dimensions cannot be considered here because our specimens preserve only the anterior portion of the body. The osteological characters concern the cephalic skeleton and include (1) skull bones devoid of ganoine, (2) parietals (= "frontals") with very narrow antorbital portion and antorbital sagittal lamina suturing to the nasal process of the premaxilla, (3) one pair of extrascapular bones, which do not reach the dorsal midline, (4) circumborbital ring open anteriorly, (5) only two suborbital bones, a small oval dorsal suborbital and a much larger ventral suborbital filling most of the area between the infraorbitals and preoperculum, (6) maxilla edentulous, (7) large quadratojugal involved in the jaw articulation, (8) four postcleithra, and (9) pectoral fins very low, joining ventrally.

The specimens from the Toarcian of Northeast Germany do as well not match with this set of characters. Even though there is agreement in characters (1), (7), and (8) they show a different development with regard to characters (2), (3), (4), (5), (6), and (9). Also, M. fegerti as well as M. lennieri have a tritoral dention.

Isanichthys Cavin and Suteethorn, 2006 (species included I. palustris Cavin and Suteethorn, 2006, I. lertboosi Deesri et al., 2014, I. latifrons (Woodward, 1893), I. luchowensis (Wang, 1974)).

Cavin and Suteethorn (2006) presented a very detailed diagnosis for Isanichthys listing the following characters: (1) skull roof bones strongly ornamented, with no continuous ganoine cover, (2) parietals (= "frontals") slightly narrower anteriorly than posteriorly, (3) ratio of parietal (= "frontal") length to postparietal (= "parietal") length less than 2.5, (4) ratio of skull length to orbit length more than 6, (5) closed orbital ring, (6) two anterior infraorbitals not in contact with the orbit, (7) anteriormost infraorbital deeper than long, (8) two supraorbitals, the anterior one generally elongated with anterior margin contacting the first and/or the second infraorbital, (9) cheek region completely covered by bones, (10) quadrate lies below the orbit, (11) preopercle slightly curved, (12) maxilla formed by a posterior rounded platelike part and a thin anterior part, which precedes the articular process, (13) one supramaxilla, (14) epiotic with a short and simple posteriorly directed process, (15) posttemporal fossa present, (16) intercalar absent, (17) basisphenoid present, (18) presence of an oral sensory canal, (19) 25 rays in caudal fin, 12 in upper lobe, and (20) 50 to 53 rows of ganoid scales along the flank and approximately 20 scales in the transverse row at the deepest level of the body.

This comprehensive diagnosis contains many primitive or poorly defined characters, e.g. characters (5) and (11) in the list above, which makes it difficult to compare our specimens to Isanichthys. However, comparison concerning characters (1), (3), and (4) reveals essential differences between the specimens from Northeast Germany and Isanichthys:

• Character (1): The skull roofing bones in our specimens are smoth and devoid of an prominent ornament.

• Character (3): The ratio of parietal (= "frontal") length to postparietal (= "parietal") length is around 3 in the two specimens from Northeast Germany instead of less than 2.5 as found in Isanichthys.

• Character (4): The ratio of skull length to orbit length is between 3 and 4 instead of more than 6.

Endocranial characters as well as numbers of caudal fin rays and vertical scale rows on the flanc cannot be specified in our specimens. Differences in characters (1), (3), and (4) are, however, substantial enough for not to include the specimens from the Toarcian of Northeast Germany in Isanichthys.

Camerichthys Bermúdez-Rochas and Poyato-Ariza, 2015 (species included C. lunae Bermúdez-Rochas and Poyato-Ariza, 2015).

The genus is monotypic and defined essentially by three unique derived characters: (1) presence of a suprapreopercular bone, (2) antorbital intermediate in depth (higher than anterior infraorbitals 1 and 2, but lower than anterior infraorbital 3), and (3) infraorbital 4 (the one forming the anterior border of the orbit) subdivided into a dorsal plus a ventral portion.

The antorbital is not preserved in the specimens from northeast Germany. Characters (1) and (3) are, however, lacking. Also, Camerichthys has a moderately tritoral dention. This, altogether, leaves no doubt that the two specimens are not congeneric with Camerichthys.

Finally, due to the relatively small size of the two specimens it might be argued that they do not represent a new genus and species but rather juvenile specimens of one of the species of Lepidotes. It is difficult to completely refute such an argument for extinct species, however, we consider it unlikely for the following reasons:

• Scale histology in reference to scale ontogeny known from modern gars shows that the two specimens are subadults or adults with fully formed bony base, dentine, and ganoine cover (Thomson and McCune 1984).

• Ontogenetic change to a tritoral dentition, characteristic of Lepidotes, is unlikely.

• Mengius acutidens gen. et sp. nov. derives from an environment clearly different from that of the Posidonia Shale.

Cladistic analysis. The cladistics analysis resulted in one most parsimonious tree (Fig. 8). Mengius acutidens gen. et sp. nov. is inferred as a member of the Lepidotidae, confirming the morphological assessment. Inside the Lepidotidae M. acutidens is sister to the species Lepidotes gigas. However, in the data matrix of Lopez-Arbarello and Sferco (2018) the coding of some character states for L. gigas was based on data from L. semiserratus, so the present result leaves the exact relationship of M. acutidens with the other Lepidotes species open, pending a revision of the genus. In total, tree stability is poor with no jackknife values over 50% found for the clade Lepidotidae or any of its subclades.

Fig. 8

Phylogenetic position of Mengius acutidens gen. et sp. nov. within the Lepidotidae. Cladistic analysis performed with TNT 1.5 (Goloboff and Catalano 2016) and based on the character matrix of López-Arbarello and Sferco (2018). Jackknife values are indicated on nodes above 50%. Data matrix for cladistic analysis available in supplementary material

Complementary fauna

The Reinberg Member of the Grimmen Formation sediments contain superabundant pseudoplanktonic “Inoceramus” dubius (Bivalvia) and Coelodiscus minutus (Gastropoda), together with a rich but taxonomically impoverished ammonite fauna, some non-belemnite gladius bearing coleoids and shrimp-like decapod crustaceans which lived in the water column (Ansorge et al. 2024). Due to the fresh water input the salinity of the restricted eastern part of the North German Basin was reduced and the exchange with the western part of the North German Basin was limited. In the Dobbertin concretions thin but distinct layers of undetermined macroalgae occur. It is unknown if these algae lived in the basin or had been washed in by rivers as was the case for the rare conchostracans (Ansorge 2003).

A diverse fish fauna with abundant Leptolepis spp. preyed on the invertebrates. Besides leptolepids, a few other rare species and individuals of Lepidotes, Dapedium, Tetragonolepis, Saurorhynchus, Grimmenichthys, unidentified "Pholidophorus", Pachycormus, Gyrosteus, and hybodontid chondrichthyans are known (e.g. Jaekel 1929, Malzahn 1963, Thies 1989b, Zessin and Krempien 2010, Konwert and Stumpf 2017, Maxwell and Stumpf 2017, Konwert and Hörning 2018, Hornung and Sachs 2020). A unique record from the Grimmen deposit is the left lower jaw of Grimmenodon aureum Stumpf et al., 2017, the only record of a pycnodontiform fish from the lower Toarcian. Supposedly, this small durophagous fish was part of the near shore biota together with benthic invertebrates, where it was caught by an unknown predator and was later regurgitated with other remains in the deeper part of the basin.

Due to the small number of identified individuals a comparison of the Grimmen Formation fish fauna with that of the Posidonia Shale is still preliminary. The restricted biota of the Grimmen Formation probably also caused a species impoverished fish fauna.

Conclusions

Two incomplete articulated specimens of fossil fish from the Lower Jurassic (lower Toarcian) Grimmen Formation of Mecklenburg and Western Pomerania (Northeast Germany) were previously identified as Lepidotes elvensis (Blainville, 1818) based on characters shown by the cranial skeleton and squamation. A restudy of the specimens revealed that they do not represent a species of Lepidotes indicated by differing dental and opercular morphology. Relationship to other lepisosteiform genera like Scheenstia, Callipurbeckia, Macrosemimimus, Isanichthys and Camerichthys can also be excluded because of differences in body and skeletal morphology and anatomy. The two specimens belong to a new genus and species, Mengius acutidens. Thin sections of scales confirm a histology similar to Lepidotes. A cladistic analysis shows that the new taxon is sister to Lepidotes. M. acutidens is unknown in the ichthyofauna of the contemporaneous lower Toarcian Posidonia Shale Formation in Northwest and South Germany. Together with other endemic fishes such as the pycnodontiform Grimmenodon aureum Stumpf et al., 2017, and the pholidophoriform Grimmenichthys ansorgei Konvert and Hörnig, 2018M. acutidens emphasises the distinct character of the lower Toarcian Grimmen Formation fish fauna of Mecklenburg and Western Pomerania.

Appendix

Remark 1.

According to Grande (2010) Ginglymodi is characterised by the following synapomorphies: (1) the anteriormost lacrimal (= anteriormost infraorbital sensu López-Arbarello (2012)) is well anterior to the orbital ring (vs. the anteriormost lacrimal is part of the orbital ring), (2) there are three or more lacrimals (= anterior infraorbitals sensu López-Arbarello (2012)) (vs. only a single lacrimal), (3) there is an autogenous splint-like quadratojugal (vs. no autogenous quadratojugal), (4) the laterally exposed portion of the subopercle has an anterodorsal process extending one third to two thirds the way up along the anterior edge of the opercle (vs. little or no projection extending dorsally), (5) the total number of principal caudal fin rays is 11–13 (usually 12) except for †semionotids and †Lepisosteus bemisi (vs. usually more or less than 12 rays), (6) there is a medial wing on the cleithrum, formed by a medial process of the cleithrum (vs. no medial wing or a wing consisting of an additional bone).

Our specimens show the derived state of characters (1) and (3). Character (2) is intermediate in that there are only the lacrimal and one infraorbital anterior to the orbit. Character (4) of Grande (2010) is also intermediate in that the anterodorsal process of the subopercle is lower than one third of the anterior edge of the opercle. Characters (5) and (6) are not observable in our specimens.

López-Arbarello (2012) proposed an alternative list of diagnostic characters for the clade Ginglymodi. It includes (in different order): (1) presence of anterior infraorbital bones, (2) premaxilla with nasal process, (3) depth of suboperculum less than half the depth of operculum, (4) gular plates absent, (5) suboperculum with well-developed, tapering dorsally ascending process, (6) closed circumborbital ring, (7) large supraorbital bones, (8) supracleithrum with a concave articular facet for articulation with the posttemporal, (9) long and narrow nasals, (10) numerous suborbital bones arranged in one row, which extends anteriorly below the orbit, (11) forward extension of the exoccipital around the vagus nerve, (12) opistothic and intercalar bones absent, (13) splint-like quadratojugal, (14) series of denticles along the ridge between the branchial and lateral surfaces of the cleithrum, (15) scale-like fin ray at the dorsal margin of the caudal fin.

López-Arbarello and Sferco (2018, 2020) have reduced and changed this list into uniquely derived synapomorphies: (1) splint-like quadratojugal, (2) anterior infraorbital bones, and homoplastic synapomorphies: (3) ectopterygoid approximately crescent shape, convex dorsally, concave ventrally, (4) antorbital bone excluded from the margin of the orbit, (5) subopercle less than half the depth of the opercle, (6) absence of gulars.

Our specimens agree with the characters (1) to (8) and (13) of López-Arbarello (2012), and with characters (1), (2), (5), and (6) of López-Arbarello and Sferco (2018, 2020). Character (9) is formed differently: the nasals are not long and narrow but broad. Interestingly, character (10) (numerous suborbital bones arranged in one row, which extends anteriorly below the orbit) displays an intermediate state. There are only four large suborbital plates in our specimens which are, however, arranged in one row anteriorly reaching below the orbit. Characters (11), (12), (14) and (15) of López-Arbarello (2012), and (3) and (4) of López-Arbarello and Sferco (2018, 2020) are not observable in our specimens.

The diagnoses of Grande (2010), López-Arbello (2012) and López-Arbarello and Sferco (2018, 2020) agree in emphasising the importance of the infraorbital bones and the suboperculum, the presence of a splint-like quadratojugal and the absence of gulars for characterising Ginglymodi. Because more than half of the characters put forward by both authors as diagnostic of the Ginglymodi occur in the specimens from the lower Toarcian of Mecklenburg and Western Pomerania, we consider them as ginglymodians.

Remark 2.

The identification of the specimen as lepisosteiform is problematical. Grande (2010) found 12 derived characters defining the Lepisosteiformes Hay, 1929: (1) the presence of opisthocoelous centra (vs. amphicoelous centra, or no solid centra), (2) the nasal process of the maxilla forms much of the ornamented dermal skull roof in the snout region (vs. a nasal process overlain by the nasals or other bones and much soft tissue, or no substancial nasal process present), (3) the lower arm of the preopercle reaches below the anterior part of the orbit (vs. not reaching that far forward), (4) supraorbital canal incorporated into the premaxilla (vs. not incorporated into the premaxilla), (5) the presence of a commissure between the right and left side supraorbital canal within the frontals (= parietals sensu Schultze 2008) (also this condition is unknown in many fossil species due to insufficient preservation), (6) the extreme posterior position of the dorsal fin, with predorsal length being 75% or more of standard length, (7) the lower jaw/quadrate articulation is positioned in advance of the orbit (vs. under or posterior to the orbit), (8) the quadrate does not articulate with the metapterygoid (vs. quadrate articulates with the metapterygoid), (9) the length of the ectopterygoid is more than twice the length of the endopterygoid (vs. less than twice its length), (10) the ectopterygoid forms the majority of the palatal region (vs. forming less than half of the palatal region), (11) the presence of a peculiar basihyal tooth plate or "tongue bone" consisting of a mosaic of bony plates or "entoglossals", (12) lateral line canal does not penetrate the posttemporal but goes directly from extrascapular to supracleithrum.

Of these derived characters only (8) is shown by our specimen. Concerning characters (2) to (5), (7), (9) and (12) they display the primitive alternative. The characters (1), (6), (10) and (11) are not observable and their state remains obscure in the specimen from North Germany.

López-Arbarello (2012) also proposed a diagnosis of the Lepisosteiformes. It contains also 12 characters but all of them are different from the characters in Grande's (2010) diagnosis: (1) ventral border of infraorbital series flexes abruptly dorsally at the anterior margin of the orbit, (2) supraorbital sensory canal marginally or not included in parietal (= postparietal sensu Schultze 2008) bones, (3) broad nasal bones, (4) closed circumborbital ring, (5) large supraorbital bones, (6) supracleithrum with a concave articular facet for articulation with the posttemporal, (7) most anterior supraorbital bone trapezoidal, longest ventrally, contacting more than one infraorbital bone, (8) three or more pairs of extrascapular bones, (9) scales with rostro-caudal articulation through anterior dorsal and ventral processes, (10) co-ossified vomers, (11) several series of denticles along the ridge between the branchial and lateral surfaces of the cleithrum, (12) scale-like ray at the dorsal margin of the caudal fin absent.

With regard to the diagnosis given by López-Arbarello (2012) our specimens agree in the characters (1) to (6). They disagree in the characters (7) to (9), which are formed differently. Characters (10) to (12) are not observable in our specimen.

López-Arbarello and Sferco (2018, 2020) revised and considerably shortened the list of López-Arbarello (2012). They accept as uniquely derived synapomorphies: (1) broad kidney shaped nasal bones, and (2) most anterior supraorbital bone trapezoidal, longest ventrally, contacting more than one infraorbital bone, and as a homoplastic synapomorphy: (3) two or more accessory rows of scales adjacent to the ventral border of the body lobe.

Of these synapomorphies only (1) applies to our specimens. (2) does not apply and (3) cannot be observed.

Because of the agreement of the specimens from North Germany in López-Arbarello's (2012) diagnostic characters (1) to (6) (which represent half of the entire number of characters) they are attributed tentatively to the order Lepisosteiformes Hay, 1929.

Remark 3.

The family name Lepidotidae Owen, 1860 was used previously to include the genera Lepidotes Agassiz, 1832, Scheenstia López-Arbarello and Sferco, 2011, Camerichthys Bermúdez-Rochas and Poyato-Ariza, 2015 and Isanichthys Cavin and Suteethorn, 2006 (e.g. López-Arbarello and Wencker 2016; López-Arbarello and Sferco 2018; Cavin et al. 2020). The autorship Owen, 1860 is, however, incorrect as shown by Koerber (2021). The correct authorship of the family name Lepidotidae is Agassiz, 1833.

The list of synonyms of Lepidotidae includes only the most important synonyms. Because of the long lasting and widespread use of the family name Semionotidae since Woodward (1890) it is far beyond the scope of this study to compile a comprehensive list of synonyms of Lepidotidae.

Remark 4.

The Lepidotidae are understood differently in López-Arbarello and Sferco (2018). In their cladogram in fig. 8 the family contains the genera Lepidotes, Camerichthys and Scheenstia with Isanichthys as the sister taxon of Lepidotidae, a view which is also expressed in López-Arbarello and Sferco (2020). In fig. 11 of López-Arbarello and Sferco (2018), however, Isanichthys is included in Lepidotidae.

The diagnostic characters proposed here for Lepidotidae are taken from López-Arbarello and Sferco (2020). These characters define node 158 in their cladogram giving rise to a clade comprising the taxa Lepidotes gigas, Scheenstia zappi, S. maximus, S. mantelli and Camerichthys lunae with Isanichthys as the sister taxon of Lepidotidae. The taxa Callipurbeckia and Macrosemimimus, also containing species traditionally considered as lepidotids are not closely related with Lepidotidae in the cladogram of López-Arbarello and Sferco (2020).