Reappraisal of the Eocene whiptail stingrays (Myliobatiformes, Dasyatidae) of the Bolca Lagerstätte, Italy

Abstract The Eocene whiptail stingrays of the family Dasyatidae from the Bolca Lagerstätte, NE Italy, are revised herein in detail. The analysis of the anatomical and morphometric features allows us to identify the species “Dasyatis” zigni (Molin, 1861) as a junior synonym of “D.” muricatus (Volta, 1796), and to assign it to the new genus Tethytrygon gen. n. This new taxon exhibits a unique combination of features (e.g., rhombic disc wider than long, elongated tail folds fail to reach the tip of the tail, thorns absent, single serrated tail sting, “caniniform” teeth on upper jaw, tooth crown ornamentation absent, 175–179 vertebrae, 108–117 pectoral radials, 24–27 pelvic radials and other features of clasper anatomy) that clearly support its attribution to the subfamily Neotrygoninae of the stingray family Dasyatidae. The morphological and phylogenetic affinities of Tethytrygon gen. n. with the living neotrygonines (Neotrygon and Taeniura) suggest a close association of this taxon with the tropical shallow‐water habitats hypothesized for the Bolca palaeoenvironment during the early Eocene. Moreover, the analysis of the fossil occurrences of the neotrygonines provides new insights into the role of the Tethys for the origin and evolutionary history of certain whiptail stingrays.


| INTRODUCTION
The family Dasyatidae, whose representatives are known as whiptail stingrays, includes about 86 living species in 19 genera of small to large stingrays (up to 2.6 m disc width, weighing up to 600 kg) living in demersal inshore habitats of continental and insular shelves up to a depth of 600 m (Last, Naylor, & Manjaji-Matsumoto, 2016;Last, White, Carvalho, et al., 2016). Morphological characters traditionally used to distinguish whiptail stingrays from other myliobatiformes include a variably depressed circular to rhombic disc not more than 1.3 times as broad as long, an angular to obtuse and sometimes very elongated snout, absent caudal and dorsal fins, greatly elongated and slender to whip-like tail, one to four long venomous spines, and a skin ranging from being completely smooth to covered with dermal denticles and thorns (Cappetta, 2012;Last, White, Carvalho, et al., 2016;Nelson, Grande, & Wilson, 2016). Although some previous studies suggest that the Dasyatidae might be a non-monophyletic group (e.g., Aschliman, Claeson, & McEachran, 2012;Carvalho, Maisey, & Grande, 2004;Lovejoy, 1996), recent morphological and molecular analyses proposed that this family should be actually regarded as monophyletic (e.g., Aschliman, Nishida, et al., 2012;Bertozzi, Lee, & Donnellan, 2016;Marramà, Klug, Vos, & Kriwet, 2018; consisting

WHIPTAIL STINGRAYS OF BOLCA
Whiptail stingrays of the family Dasyatidae are among the first cartilaginous fishes described from the Bolca Lagerstätte, but their taxonomic history is characterized by an intricate and complex scenario. The first taxon was described and figured by Volta (1796, pl. 9, fig. 1) under the name Raja muricata based on a single specimen from the Gazola collection in part and counterpart, today housed in the Museum National d'Histoire Naturelle in Paris, registered under MNHN F.Bol.564 ( Figure  1). On the same plate, Volta (1796, pl. 9, fig. 2) figured an incomplete portion of a tail preserving a caudal sting using the same name that was later assigned to Taeniura knerii by Molin (1861), a taxon considered a synonym of Urolophus crassicaudatus by Eastman (1904Eastman ( , 1905aEastman ( , 1905b. Later, de Blainville (1818) assigned MNHN F.Bol.564 to Trygonobatus vulgaris, and Agassiz (1835), Agassiz (1833Agassiz ( -1844 to Trygon gazzolae, without any reliable new description or anatomical interpretation. Subsequently, based on specimen MGP-PD 159Z/160Z, Molin (1861) created the genus Alexandrinum, which was specified as Alexandrinum molini by de Zigno (1874a) and figured in unpublished material ( Figure 2). All these taxa and their respective specimens were referred to Trygon by Jaekel (1894), with R. muricata being considered as the holotype (Trygon Cuvier, 1816 is today regarded as a junior synonym of Dasyatis Rafinesque, 1810). The synonymy of Trygon muricatus was subsequently confirmed by Eastman (1904Eastman ( , 1905aEastman ( , 1905bEastman ( , 1911. (Volta, 1796)  Based on another single incomplete specimen lacking part of the tail and the sting (MGP-PD 150Z/151Z), Molin (1861) described another dasyatid taxon as Anacanthus zigni. As reported later by de Zigno (1874a, 1874b), who also figured the specimen in unpublished material ( Figure  3), Molin diagnosed and distinguished the new taxon from F I G U R E 2 (a-c) Tethytrygon muricatus (Volta, 1796)  R. muricata based on the supposed absence of a caudal sting and a tail that is shorter than the disc length. Subsequently, Jaekel (1894) recognized that the presence of a very short tail and the absence of a sting in MGP-PD 150Z/151Z were due to the incompleteness of the specimen, and assigned the species to the genus Trygon. However, he maintained and distinguished Trygon zigni (Molin, 1861) from T. muricatus (Volta, 1796) based on the "much smaller size […] a less rounded outline of the disc […] and pelvic fins triangular in shape." Since then, no detailed anatomical descriptions and taxonomic interpretation of the whiptail stingrays of Bolca were carried out. Our revision of the material showed that no substantial morphological differences support the hypothesis that MGP-PD 150Z/151Z (nor MGP-PD 159Z/160Z) should be recognized as a different species. The analysis of the anatomical and morphometric features allow us to recognize the species "Dasyatis" zigni (Molin, 1861) as a junior synonym of "Dasyatis" muricatus (Volta, 1796) and to assign it to the new genus Tethytrygon gen. n. The new taxon is represented by 13 partially complete and articulated skeletons (Figures 1-4). The large number of available specimens and their good preservation allowed for the recognition and description of several skeletal and dental characters, which are useful to distinguish and separate the taxon from any other known living and fossil dasyatid (see the detailed anatomical description in the Supporting information Appendix S1). The specimens examined comprise different ontogenetic stages, with the largest one (an adult male) being characterized by 60 cm disc width and possibly reaching 150 cm in total length. The disc of Tethytrygon gen. n. is rhombic, not wing-like, reaching the maximum width in the anterior third of disc length. The disc length is slightly shorter than the disc width (0.9 times), whereas the total length is about 2.6 and 2.8 times those of the disc width and disc length, respectively. The tail is long and about 1.8 times the disc width. Tethytrygon muricatus lacks dorsal fins, whereas a single serrated sting can be recognized in most of the specimens. The placement within the subfamily Neotrygoninae is particularly supported by the presence of files of "caniniform" teeth in the upper jaw ( Figure 5), which represent a unique and derived trait for Neotrygon and Taeniura among stingrays (Cappetta, 2012;Last, Naylor, & Manjaji-Matsumoto, 2016;Last, White, Carvalho, et al., 2016) and supports the grouping of Tethytrygon gen. n. with these genera in our phylogeny. Tethytrygon muricatus is a unique neotrygonine in having the following autapomorphic traits: large size (up to 60 cm DW and possibly 150 cm TL), long tail (170.4%-184.7% DW) and low number of monospondylous trunk vertebrae (23-26). Additionally, Tethytrygon gen. nov is also characterized by a unique combination of morphological and meristic characters that allow to distinguish it from the other neotrygonines (Table 1). These features include a disc rhombic in shape, disc length 87.2%-95.2% DW, total length 249.5%-263.0% DW, subtriangular pelvic fins 24.6%-29.9% DW and eye diameter 2.7%-4.5% DW. The skin of T. muricatus is mostly smooth without thorns but with small scattered star-shaped dermal denticles in largest individuals. A single serrated sting of 26.1%-32.7% DW can be recognized in most of the specimens. The vertebral column is composed of 175-179 vertebrae. The pectoral disc contains 108-117 pectoral radials of which 49-53 are propterygial, 16-20 are mesopterygial, and F I G U R E 5 (a-j) Tethytrygon muricatus (Volta, 1796) from the Eocene of Bolca Lagerstätte (specimen MCSNV IG.186653) . (a-e) A single tooth from the file of "caniniform" teeth in (a) occlusal, (b) lingual, (c) labial, (d) lateral and (e) basal view; the picture depicted in A1 represent the same tooth still in place (the cusp has broken during the extraction). (f-j) Another isolated tooth (not coming from the file of "caniniform" teeth) in (f) occlusal, (g) lingual, (h) labial, (i) lateral and (j) basal view. Scale bar is 500 μm [Colour figure can be viewed at wileyonlinelibrary.com] T A B L E 1 Morphological and meristic characters useful to distinguish Tethytrygon gen. n. from the living neotrygonines Taeniura and Neotrygon. All measurements as percentage of disc width (%DW) and mean values are within parentheses. The living species include Neotrygon annotata, N. australiae, N. caeruleopunctata, N. indica, N. kuhli, N. leylandi, N. ningalooensis, N. orientalis, N. picta, N. trigonoides, N. varidens, Taeniura lymma andT. lessoni. Data from Schwartz (2005), Schwartz, (2007), Schwartz (2008), Last and White (2008), Last, White, Carvalho, et al. (2016), ,  and Pavan-Kumar, Kumar, Pitale, Shen, and Borsa (2018)

Morphometric character Tethytrygon Taeniura Neotrygon
Max disc width ( 40-45 are metapterygial radials. The pelvic fins have 24-26 radials. The tail folds are located posterior to caudal sting origin and fail to reach the tip of the tail as in Neotrygon. The teeth are rhombic in occlusal view and possess well-marked concavely arched cutting edges. Their lingual surface is low and strongly concave and the crown ornamentation is absent. Carvalho et al. (2004, fig. 51) tentatively placed T. muricatus (as "Dasyatis" muricata) within the Myliobatiformes. Though they recognized its affinities with dasyatids, they placed it conservatively in an unsolved polytomy together with Dasyatis, Himantura, Styracura (as "Himantura"), Pteroplatytrygon and Taeniura. Our analysis of 103 traits coded for 30 taxa produced a single parsimonious tree (length 216 steps, C.I 0.65, R.I 0.79) that resolved many of the systematic affinities of T. muricatus ( Figure 6). A complete list of synapomorphies at each node is listed in The absence of ribs (ch. 80), traditionally recognized as a synapomorphy of stingrays, does not support the clade in our phylogeny, since skates (here represented by Raja) lack ribs as well (Marramà, Schultz, & Kriwet, 2018). Hexatrygon (Sixgill stingray) is inferred to be the sister to all other stingrays in most analyses based on morphological data (Aschliman, Claeson, et al., 2012;Aschliman, Nishida, et al., 2012;Carvalho et al., 2004;Claeson et al., 2010;, but not in recent molecular phylogenies, where it was recovered nested within myliobatiformes being sister to the urolophids (Naylor, Caira, Jensen, Rosana, Straube et al., 2012;Naylor, Caira, Jensen, Rosana, White et al., 2012) or to Gymnura (Bertozzi et al., 2016).

| PHYLOGENETIC ANALYSIS
Our phylogeny detected a dichotomous nature of remaining myliobatiformes as determinated by . The dichotomy is formed by two main clades that correspond in part to the superfamilies Myliobatoidea and Dasyatoidea. The nature of the dichotomy is possibly linked to the different calcifications of radial cartilages, body shapes and swimming modes detected in these two main groups by Schaefer and Summers (2005). The monophyly of the myliobatoids, including Gymnura as sister to pelagic stingrays, is F I G U R E 6 The single parsimonious tree retrieved in tnt v.1.5 based on 103 morphological characters showing the hypothetical relationships of Tethytrygon muricatus (Volta, 1796) within the Myliobatiformes. Numbers on nodes indicate the Bremer support. Extinct taxa are marked with a dagger. The list of synapomorphies on each node (capital letters) is given in Table 2 supported herein by seven synapomorphies: short orbital region with more anteriorly placed supraorbital and postorbital process (ch. 10[1]); mesopterygium fragmented (ch. 27[1]); lateral expansion of radials in pectoral region (ch. 28[1]); caudal fin absent (ch. 34[2]); first segment of propterygium adjacent to anterior margin of antorbital cartilage or anterior to margin of nasal capsule (ch. 81[3]); "crustal" calcification pattern of radials (ch. 100[0]); and wing-like body shape with pectoral fins greatly expanded (ch. 101[1]). This clade includes those stingrays with crustal calcification of radials and a wing-like body shape that possibly reflect their unique oscillatory swimming mode (Schaefer & Summers, 2005).
Recent molecular analyses resolved Gymnura as sister to Urolophus (Aschliman, Nishida, et al., 2012), Plesiobatis , Hexatrygon (Bertozzi et al., 2016), or placed it much closer to the base of all myliobatiformes (Last, White, Carvalho, et al., 2016). The family Myliobatidae (including Aetomylaeus, Myliobatis, Aetobatus, Rhinoptera, Manta and Mobula) is herein detected as monophyletic and well supported (Bremer value 5) by 28 characters (see Table  2). The monophyly of the clade Dasyatoidea (including all remaining stingrays) is weakly supported (Bremer value 1) by a single character, the spiracularis split into lateral and medial bundles, with the medial bundle inserting on to the posterior surface of Meckel's cartilage and the lateral bundle inserting onto the dorsal edge of the hyomandibula (ch. 88 [1]). This group includes stingrays having rhomboidal or oval disc shapes and "catenated" calcification of radials, which reflect their undulatory swimming mode and benthic habits (Schaefer & Summers, 2005). The family Urolophidae (Urolophus + Trygonoptera) is sister to all dasyatoids, and its   Figure 8) the systematic position of Plesiobatis, Asterotrygon and Heliobatis was poorly resolved and the analysis detected two different hypotheses. Possibly due to the recoding of some characters (Supporting information Appendix S1) and to the inclusion of the new taxon described herein, our new analysis detected a single tree in which Plesiobatis is more basal than the Eocene freshwater stingrays Asterotrygon and Heliobatis. These two fossil taxa form a monophyletic clade supported by a single character (caudal fin reduced to tail folds; ch. 34[1]). In fact, since this character is absent in the outgroups, Hexatrygon, urolophids and Plesiobatis (all of them having a fully developed caudal fin), the reduction of the caudal fin to tail folds seems to have been achieved originally in the common ancestor of Asterotrygon and Heliobatis and later, independently, in more advanced dasyatids. Although the relationship between Asterotrygon and Heliobatis is weakly supported (Bremer value 1), this might corroborate the hypothesis that the two genera diverged after their common ancestor invaded the freshwater system of Green River Formation, contrary to the hypothesis of Carvalho et al. (2004), who hypothesized that Asterotrygon and Heliobatis might have invaded independently the Eocene freshwaters of Fossil Lake. It is therefore reasonable to recognize a single monophyletic family, which includes these two extinct genera (Heliobatidae Marsh, 1877 ). Although a close relationship between the freshwater potamotrygonids and the marine stingray Styracura is almost certainly true as detected by morphological, molecular and chrono/geographic evidences (Aschliman, Claeson, et al., 2012;Bertozzi et al., 2016;Carvalho et al., 2004;Carvalho, Loboda, & Silva, 2016;Lovejoy, 1996;Naylor, Caira, Jensen, Rosana, Straube et al., 2012;Naylor, Caira, Jensen, Rosana, White et al., 2012), our phylogeny did not recognize Styracura as a genuine member of the family Potamotrygonidae, due to the fact that the Styracura lacks some characters of the lateral-line, and pectoral and pelvic fin skeleton typically found in freshwater potamotrygonids . The monophyletic status of whiptail stingrays of the family Dasyatidae (including here Himantura, Neotrygon, Taeniura, Pteroplatytrygon, Pastinachus and Dasyatis), as recognized by Aschliman, Nishida, et al. (2012), , Bertozzi et al. (2016), , but not by Carvalho et al. (2004), Aschliman, Claeson, et al. (2012) and Lim et al. (2015), is recognized and supported herein by two features: ventral terminal cartilage free of axial cartilage (ch. 87 [1]) and presence of sexual heterodonty (ch. 89[1]). The presence of tail folds used to diagnose the family Dasyatidae by Bigelow and Schroeder (1953), Compagno andRoberts (1982, 1984) and Nishida (1990) is not supportive of the clade because they are also present in Styracura, some freshwater potamotrygonids and extinct heliobatids. The sister-group relationship between the urogymnines Protohimantura and Himantura is again recognized as in  supported by a second transverse tooth keel in these taxa (ch. 99[1]), and mid-dorsal surface of disc covered by heartshaped denticles arranged in an antero-posteriorly directed patch with sharply defined outlines (ch. 102[1]). The placement of urogymnines as the sister of all other dasyatids is in accordance with molecular analysis presented by Puckridge, Last, White, and Andreakis (2013) but inconsistent with the molecular and morphological phylogenetic results of Lim et al. (2015) and . Tethytrygon gen. n. is clearly a genuine member of the subfamily Neotrygoninae (including the living Neotrygon and Taeniura) which is supported herein by three synapomorphies: spiracularis muscle projecting ventrally and posteriorly beyond hyomandibulae and both sets of jaws to insert dorsal to coracomandibularis (ch. 36[1]), presence of anterior process of the Meckel's cartilage (ch. 92[1]), and file of "caniniform" teeth in the upper jaw (ch. 103[1]). It is interesting to note that the monophyly of living neotrygonines has been also detected by the molecular analyses of Aschliman, Nishida, et al. (2012), Puckridge et al. (2013), Lim et al. (2015) Bertozzi et al. (2016) and , as well as in the morphology-based study by . Although Tethytrygon gen. n., Neotrygon and Taeniura have been resolved in a polytomy, the analysis detected some autapomorphic characters (not shown) useful to distinguish the three genera. For example, Tethytrygon gen. n. is unique in the absence of thorns (ch. 69[1]), which instead are present in at least a single antero-posteriorly directed row of thorns dorsally on disc in living neotrygonine genera (e.g., Last, White, Carvalho, et al., 2016), whereas Neotrygon can be distinguished from Taeniura in the absence of levator and depressor rostri muscles (ch. 66[0]), and in the presence of a jugal arch (ch. 75[0]). Finally the clade Pteroplatytrygon + (Dasyatis + Pastinachus) is recovered here as the most derived clade among dasyatoids, supported by two synapomorphies: pseudosiphon present in claspers (ch. 82[1]) and spiracularis split into lateral and medial bundles, with the medial bundle inserting onto the posterior surface of Meckel's cartilage and the lateral bundle inserting onto the dorsal edge of the hyomandibula (ch . 88[1]). The bootstrap tree (Figure 7) looses resolution as expected, but a close relationship between Tethytrygon, Neotrygon and Taeniura (Neotrygoninae) is still retrieved.

| Comparison and relationships
The detailed morphological analysis of T. muricatus (Supporting information Appendix S1), has revealed the presence of a number of characters that strongly support its inclusion within the order Myliobatiformes, including the absence of rostral cartilage, presence of a broad and shelflike postorbital process, thoracolumbar synarcual, serrated tail sting and basihyal separated from first hypobranchials (e.g., Compagno, 1977;Carvalho et al., 2004;Aschliman, Claeson, et al., 2012). The placement of T. muricatus within the Dasyatidae is supported by the ventral terminal cartilage that is free of the axial cartilage, and presence of sexual heterodonty. Moreover, a combination of several plesiomorphic characters argues against the placement of T. muricatus in other clades. For example, the presence of tail folds excludes its assignment to myliobatids and dasyatoids characterized by developed caudal fin (e.g., urolophids and urobatis). The absence of angular and secondary cartilages separate the new genus from potamotrygonids, whereas the first segment of the propterygium adjacent to the anterior margin of the antorbital cartilage or anterior to the margin of the nasal capsule separate T. muricatus from non-dasyatids dasyatoids (posterior to the mouth, between mouth and antorbital cartilage, or adjacent to the nasal capsule in these latter). An external margin of the mesopterygium that is more or less straight and not fused to radials exclude any relationship between T. muricatus and Gymnura (undulated, not fused to radials) or the Urolophidae (highly sinuous, fused with radials; e.g., Carvalho et al., 2004). Moreover, the absence of all the shared derived traits characterizing Gymnura and pelagic stingrays (Table 2) supports the exclusion of Tethytrygon gen. n. from the group of myliobatoid stingrays.
The morphological and phylogenetic analysis identified Tethytrygon gen. n. as a genuine member of the Neotrygoninae, in a polytomous relationship with the extant representatives of this subfamily Neotrygon and Taeniura. The placement within the subfamily is supported by the presence of the anterior processes of the Meckel's cartilage and the file of "caniniform" teeth in the upper jaw. Tethytrygon gen. n. differs from the two living neotrygonine genera by its larger size, longer tail and lower number of trunk vertebrae in addition to morphometric and meristic features (Table 1). Tethytrygon gen. n. can be readily separated from Taeniura by proportional measurements in disc length, snout to pectoral-fin insertion, preorbital and preoral length, eye diameter, snout to maximum disc width, and tail length (Table 1). Moreover, the absence of any tooth ornamentation and thorns, and tail folds failing to reach the tip of the tail in Tethytrygon gen. n. distinguish it furthermore from Taeniura. Tethytrygon gen. n. differs from Neotrygon in having different proportions of the snout and orbit to pectoral-fin insertions, pectoral-fin insertion to sting length, eye diameter, pelvic fin, sting, tail and total lengths, and vertebral counts (Table 1).

| Palaeoecology, palaeobiogeography and evolutionary significance
Extant stingrays of the subfamily Neotrygoninae are demersal, benthic marine batoids occurring inshore on continental or insular shelves at depths up to 90 m (Last, White, Carvalho, et al., 2016;Nelson et al., 2016). Neotrygon and Taeniura mainly inhabit warm-temperate and tropical shallow waters, and are often associated with the coral reefs of the Indian Ocean, and Indo-Australian Archipelago, feeding mainly on small bony fishes, crustaceans, worms and bivalves (Last, White, Carvalho, et al., 2016). In this perspective, the presence of several specimens of Tethytrygon gen. n., which represents the most common batoid in the Bolca palaeobiotope, suggests close affinities of this taxon with the shallow-water habitats, possibly associated with coral reefs, hypothesized for the Pesciara palaeobiotope (Marramà, Bannikov, Tyler, Zorzin, & Carnevale, 2016;Papazzoni & Trevisani, 2006).
Although the fossil record of Dasyatidae is extensive and well documented, probably dating back at least to the Early Cretaceous (Cappetta, 2012;Underwood et al., 1999), fossils of the subfamily Neotrygoninae are rare and, with the exception of Tethytrygon gen. n., solely represented by isolated teeth. However, the paucity of the fossils probably represents an artefact, since many neotrygonine teeth might have been misassigned to Dasyatis, which has been traditionally used as catch-all genus for many fossil teeth exhibiting "dasyatoid" morphology (Cappetta, 2012;Underwood et al., 1999). Fossils of the genus Neotrygon have been reported so far only from the middle to late Eocene deposits of the Fayum area, Egypt. The single tooth figured by Underwood et al. (2011, fig . 7p) is very similar to teeth of Tethytrygon gen. nov. Based on palaeobiogeographic, palaeoecological and palaeoenvironmental evidences, we do not exclude that teeth reported as Neotrygon sp. by Underwood et al. (2011) may belong to Tethytrygon gen. n. Conversely, the genus Taeniura was reported from several localities. However, the relative abundance of Taeniura in the fossil record from the Miocene to the Pliocene might be an artefact since teeth traditionally reported as Taeniura grabata and Taeniura cavernosa should be referred to the dasyatid genus Taeniurops (subfamily Dasyatinae), recently resurrected by Last and Stevens (2009) based on unambiguous morphological and dental differences with respect to Taeniura (see Cappetta, 2012). Thus, reliable reports of Taeniura (as T. sp.) appear to be solely restricted from the middle to late Eocene of the Fayum area, Egypt (Underwood et al., 2011). Teeth of Taeniura sp. are also reported from the lower Miocene of Brazil (Aguilera et al., 2017) and from the Pliocene of Libya (Pawellek et al., 2012), although it is not clear whether the authors recognized their affinities with Taeniurops (T. grabata or T. cavernosa) or Taeniura (T. lymma or T. lessoni). Therefore, the oldest remains referable to neotrygonines are from Eocene tropical shallow Tethyan localities (Bolca and Fayum area; Figure 8).
The Bolca chondrichthyan assemblage is remarkably different from those of other contemporaneous Boreal (London Clay, Paris basin, Lede Sand Formation, Fürstenau Formation, Lillebaelt Clay) or Tethyan (SW France and Northern Morocco) deposits, suggesting that its taxonomic composition is largely influenced by the different palaeoenvironmental setting . Conversely, the Bolca palaeoenvironmental and palaeoecological characters appear to be more consistent with the tropical shallow settings reported from south-western Morocco and, even more, with those of the Fayum area in Egypt . Like the latter, in particular, the Bolca fauna is characterized by the presence of small odontaspidids, small carcharhinids, and juvenile triakids, all generalist feeders preying on small nectobenthic preys and zooplanktivorous coastal bony fishes. Among batoids, the Fayum area and the Bolca Lagerstätte share the presence of thornbacks (Platyrhinidae) and, as detected in the present study, neotrygonines, which are absent in other deposits Underwood et al., 2011).
Today, the neotrygonines Neotrygon and Taeniura are restricted to continental and insular shelves of the Indian Ocean and Indo-Australian Archipelago (Last, White, Carvalho, et al., 2016). Divergence time estimates indicate that living neotrygonines diverged from other dasyatids in the Late Cretaceous and that Neotrygon diverged from Taeniura around the K-Pg boundary (Puckridge et al., 2013). However, Aschliman, Nishida, et al. (2012) placed the divergence of neotrygonines from Dasyatis around 50 million years ago. Later, a series of rapid cladogenetic events (triggered by tectonics and eustatism) were probably responsible for the isolation and high diversity of Neotrygon species in the Indo-West Pacific area (Puckridge et al., 2013).
The authors also suggested an austral origin for the genus Neotrygon. Although collecting and taphonomic biases must be considered, since the earliest known neotrygonines appear to be the Ypresian to Priabonian occurrences of Bolca and Fayum area, one can suppose a Tethyan origin for the group and an eastward migration of its representatives from the Tethys during the Eocene, to the Arabian Peninsula and the Indo-Australian Archipelago in the Miocene, following the shift of the centre of marine biodiversity across the globe from the Eocene to today (Renema et al., 2008). This pattern was also highlighted at least for two other dasyatid subfamilies, the Hypolophinae and the Urogymninae, whose more abundant fossil record indicates an Eocene origination in the Tethys, followed by a widespread colonization of the proto-Mediterranean Sea and Indo-Pacific from late Palaeogene to the early Neogene (Adnet et al., 2018;.

| CONCLUSIONS
The revision of the Eocene stingrays from the Bolca Lagerstätte traditionally referred to "Dasyatis" muricatus and "D." zigni allowed a detailed reinterpretation of their morphology and taxonomic status. A unique combination of morphological features allowed the recognition of a new genus of the family Dasyatidae, Tethytrygon gen. n. The phylogenetic analysis suggested close affinity to the living representatives of the subfamily Neotrygoninae. The scarce fossil record of neotrygonines seems to suggest a Tethyan origin for the group, and that their modern distribution restricted to the Indian Ocean and Indo-Australian Archipelago may be the final result of their spatial dynamics across the Palaeogene and Neogene, following the eastward shift of the marine centre of palaeobiodiversity across the globe, a model also detected for hypolophines and urogymnines, among stingrays. two unpublished plates of Bolca fossil rays by Achille De Zigno housed at the Library of the Department of Geosciences of Padova University. Luca Giusberti (Padova University) is thanked for the useful suggestions and support in this. The authors also thank Faviel Lopez Romero (University of Vienna) for having cleared and stained the Neotrygon and Taeniura specimens; Cathrin Pfaff (University of Vienna) in performing micro-CT scans of Taeniura used for comparisons; and Charlie Underwood for the useful discussion about the Eocene neotrygonine teeth from the Fayum area in Egypt, and for kindly providing pictures of Neotrygon kuhlii (BMNH 2015.1.25.6) and the Fayum tooth used for comparisons. The manuscript was also greatly improved with the constructive reviews provided by Todd Cook (Penn State Behrend) and by editorial comments.