A Flying Fish Across the Tethys: Thoracopterus wushaensis Tintori et al., 2012, from the Pelsa/Vazzoler Fossil-Lagerstätte (Late Ladinian, Middle Triassic) in the Dolomites (Italy)

Abstract

For the first time, the same fish species is recorded from Triassic rocks at both ends of the Tethys, namely, Thoracopterus wushaensis Tintori et al., 2012. First described from the southwestern China Xingyi Fauna (late Ladinian, Middle Triassic), it is now also known from a coeval site in the Dolomites in northern Italy. T. wushaensis is considered a flying fish, owing to its elongated pectoral and pelvic fins, tail with a ventral lobe much larger than the dorsal one, and peculiar enlarged neural spines in the caudal region. Taxonomical problems among Peltopleuriformes are also discussed on the basis of an improved comparison between eastern and western Tethys fish assemblages. This finding bolsters the marine vertebrate similarity between China and Europe during the Middle Triassic. Furthermore, because T. wushaensis from the Pelsa/Vazzoler is part of a very rich assemblage yielding several fishes, and also dozens of invertebrate taxa, it also supports the hypothesis of a global bioevent among marine fauna during the late Ladinian, well before the San Cassiano Fauna or the Carnian Pluvial Episode.

1. Introduction

Fossil flying fishes have been known since 1858, when Bronn described the first species (Thoracopterus niederristi Bronn, 1858 [1]) from the early Carnian (Late Triassic) of Raibl (Cave del Predil in Italian, Udine, Italy), which then belonged to Carinthia in the Austro-Hungarian Empire [2]. More recent descriptions of this species are by Griffith [3] and Lehman [4], although from a different site (Polzberg, Austria), and their anatomical interpretation is partly questioned by Tintori and Sassi [5]. However, recently, Tintori and Conedera (submitted) stated that the Polzberg material is different from the Raibl specimens and erected a new species for the Polzberg Thoracopterus. Abel [6] also included Gigantopterus telleri Abel, 1906, from Polzberg in this group of flying fishes; the latter material has also been redescribed by Griffith [3], who noticed the great anatomical similarity between this species and the type species of Thoracopterus, even though the preservation of the Polzberg material is extremely poor (pers. obs. in 1988, 2013, 2025). The Polzberg fauna is now considered as late early Carnian [7], thus somewhat younger than the “original” Raibl assemblage yielding T. niederristi [8]. Griffith [3] also erected the family Thoracopteridae, considering it as a member of Perleidiformes. Tintori and Sassi [9] added T. magnificus from the Norian (Late Triassic) of Lombardy (northern Italy), and soon after, the same authors erected T. martinisi Tintori and Sassi, 1992, again from the Norian but in Carnia (Friuli-Venezia Giulia, northeastern Italy). In the same paper, Tintori and Sassi [5] moved G. telleri to Thoracopterus owing to the very similar skull anatomy and considered this genus to be related to Peltopleurus (thus belonging to Peltopleuriformes) rather than Perleidus (Perleidiformes). Thus, for about 150 years, Triassic flying fishes have been known only from the Late Triassic (early Carnian to middle/late Norian) of the Western Tethys, whereas other supposed “flying fishes” proved not to be able to “fly” due to a lack of characteristics related to flying, other than elongated/enlarged pectoral and possibly pelvic fins [5]. Fossil flying fishes of the genus Thoracopterus, as well as the modern Exocoetidae, were able to lift themselves out of the water with great speed by means of their propeller, the powerful caudal region where deeply modified (longitudinally enlarged) neural spines are related to a very stout ventral lobe of the caudal fin in both fossil and recent taxa [5] (Figures 10 and 11).

Recently, the discovery of several very rich Middle to early Late Triassic fish and marine reptile faunas in southern China has allowed expansion of the geographical and stratigraphical distribution of many marine vertebrate groups down to genus level, including fishes and many others [10,11,12]. The similarity between the Triassic fish assemblages at the two ends of the Tethys was first noted by Tintori and Lombardo [13], and since then, several taxa from the new Chinese faunas can be easily related to well-known ones from classical localities in Europe, especially those from the Southern Calcareous Alps such as the Monte San Giorgio, located across the Italy/Switzerland boundary just north of Milan, as well as other Middle Triassic (Perledo-Varenna, Lake Como, Lecco, Lombardy, Italy) or early Carnian Raibl/Cave del Predil (Friuli-Venezia Giulia) faunas [2].

Among the new Chinese assemblages, the late Ladinian Xingyi Fauna has proved to be particularly interesting for actinopterygian fishes because it yields, among many other taxa, the oldest pholidophorids [14] as well as the oldest Thoracopterus species, T. wushaensis [15]. In both cases, the previous oldest records were from the early Carnian Raibl/Cave del Predil site, respectively, Knerichthys bronni (Kner, 1866) [16,17] and T. niederristi itself. Chinese specimens have been collected by an international team from Peking University (PKU; Beijing, China) during a very large bed-by-bed excavation [18] near the village of Nimaigu (Wusha District, Xingyi City, Guizhou Province, China), where the pholidophorid Malingichthys was erected based on two species from several specimens from bed 35 [14], while T. wushaensis is from beds 42–45 [15]. Although it is almost certain that there are no flying fishes below bed 42, it is possible that Malingichthys is also present in the upper beds (above bed 40), as most of the small specimens from those beds have not been prepared yet (pers. obs.). The fish assemblage of the Xingyi Fauna comprises many different small peltopleuriforms that are common all along the vertebrate level, in both the Lower (below bed 40) and Upper horizons (but with the very small Habroichthys orientalis (Su, 1959) [19] dominating the Lower Horizon and apparently absent in the Upper Horizon), together with the mid-sized halecomorph Asialepidotus shingyiensis Su, 1959 [19,20]. H. orientalis is often preserved in mass mortality surfaces, some of them in bed 35, which has about 20 specimens per square meter (pers. obs. during excavation in 2013, Tintori et al., submitted). Also quite common in the Upper Horizon is an as-yet undescribed species of Marcopoloichthys, a small fish widespread across the Tethys and most of the Triassic [21,22]. Thus, the fish assemblages from both the Lower and the Upper Horizon show a clear Middle Triassic mark, also because of the very small size of several taxa (“miniature fossil fishes” in [14]), but with the addition of new groups such as the Pholidophoridae and the Thoracopteridae that would later thrive in the Late Triassic.

This marine vertebrate assemblage (fishes + marine reptiles) also proves to be very important because there has been good dating through ammonoids [23] that points to a middle late Ladinian age, although most of the ammonoid taxa are endemic for southern China. Moreover, the Xingyi Fauna is considered a key point in the achievement of a more oceanic way of life for marine vertebrates [18].

A few years ago, a new late Ladinian Fossil-Lagerstätte was discovered in the Belluno Dolomites (Figure 1), just south of Mt. Civetta, near the Casera Pelsa and the Vazzoler Lodge [24], where in 1975 the author found his first remains of a Triassic fish, a fragment that many years later proved to belong to Marcopoloichthys [21,25]. Although the fieldwork has been just preliminary, a fish assemblage of at least 15 taxa has been recovered, mostly from a single bed (Site A, bed 96). This major fish bed is intercalated among three layers that yield a very rich, silicified benthic invertebrate fauna [26,27,28], with other layers also being rich in terrestrial plant remains. Some ammonoids allowed precise dating to the A. neumayri Subzone in the late Ladinian (Longobardian) of the Tethys area (P. Mietto, 2017, written comm.); thus, it is comparable to the Chinese Xingyi Fauna [23]. The very similar age of the two fossil assemblages is now further supported by the presence of T. wushaensis in the Pelsa/Vazzoler Fossil-Lagerstätte as well as in the type area of Xingyi. It is the first time that the same fish species can be identified in both ends of the Triassic Tethys.

2. The Fish Fauna from the Pelsa/Vazzoler Fossil-Lagerstätte

The Pelsa/Vazzoler fossil site represents a quite small (just a few square km), inner lagoon–intraplatform basin surrounded by the Middle Triassic Civetta-Pale di San Martino carbonate platform, one of the largest of the Dolomites [24,29,30]. Invertebrates found just above and below the fish bed are clearly allochthonous, reworked from the basin rim [27], and they represent small coral reefs with plenty of mollusks, echinoderms, corals, brachiopods, and sponges, as well as small incertae sedis organisms. The benthic invertebrate assemblage is very similar to the famous early Carnian San Cassiano Fauna [27], which is somewhat younger than the Pelsa/Vazzoler Biota, and is considered the most diverse invertebrate assemblage before the Pliocene [31]. We must keep in mind that for the San Cassiano Formation, collecting started around the middle of the 19th century and is still active in several localities of the Dolomites, whereas for Pelsa/Vazzoler, most specimens are from a single small outcrop with just three beds yielding the silicified invertebrates. To date, more than 150 benthic invertebrate species have been detected after freeing isolated specimens by acid treatment of around 300 kg of limestone blocks [27], compared to more than 1500 nominal species for the San Cassiano Fauna [31]. However, in the latter fauna, vertebrates are very rare, and the very scarce fish remains are just a few scattered teeth [32] from durophagous sharks and possibly Colobodus, while here we are dealing with complete and well-articulated fish specimens, the majority of which are of very small size (miniature fossil fishes < 40 mm).

Most of the identifiable fishes (mostly complete) recovered so far from this new site are from a single bed (Site A, bed 96) from a small outcrop near the Casera Pelsa, in front of Torre Venezia Peak (Venice Tower) at the southwest of Mt. Civetta, but also from other small outcrops around the main site. Most specimens are very small, less than 40 mm long, with Habroichthys averaging around 20–25 mm in standard length [33]. This “miniature fossil fish” assemblage is quite characteristic of the Triassic Middle Fish Fauna [14].

The main fossiliferous site (Site A, Figure 2) is a small artificial excavation opened by the landowners around 25 years ago. In 2016, the author, together with his colleague M. Montagna, was trying to find the site where, in 1975, he discovered a fragment of Marcopoloichthys [21,25] and one fossil insect. Although the original site was not found, new outcrops proved to be very rich, especially the small, old excavation near the Casera Pelsa (“casera”: a traditional summer hut for cheese production in the Italian Alps) at about 1.800 m a.s.l., which yields a large array of vertebrates, invertebrates, and plants. Here, a section of about 1.5 m was irregularly outcropping, showing three black limestone beds with silicified shallow water invertebrates (beds 98, 97, and 93 from the top down), while a laminated limestone (bed 96) yielded fishes, plants, rare insect remains, and the bivalve Bositra wengensis (Wissmann, 1841) [34] in the lower part. The very base of this “fish bed” shows extensive bioturbation and appears pinkish at the top of the underlying bed due to the presence of volcanic ash. Bed 96 is very broken, so usually only slabs less than 15–20 cm can be collected. This is due to the very superficial position of the beds in an alpine pasture, where, especially in winter, low temperatures can freeze the thin overlying soil and superficial beds, causing them to fracture and fragments to be displaced.

Two very short excavations were carried out in 2017 and 2018 to preliminarily exploit the site, while in 2024, a larger excavation campaign was performed. During the first collection, a large amount of blocks yielding silicified invertebrates was collected and dissolved by acid. Several thousand specimens, mostly very small, have been selected, and their study is still underway [27].

Although only a very small surface of bed 96 (now about ten square meters) has been investigated, at least 15 fish taxa have been recognized, most of them actinopterygians, but with sharks and coelacanths also present. The assemblage is dominated by Habroichthys and Marcopoloichthys [33], which together make up around 60% of the specimens collected. Placopleurus (Conedera et al., submitted) and Peltopleurus are also common (totalling about 16%), whereas Saurichthys, shark, and coelacanth remains are rarer, and about 20% of the specimens still need to be identified but show quite an interesting diversity. A single specimen of T. wushaensis was collected from the scattered scree of bed 96 when the site was discovered in September 2016.

Sharks are scarce in most Middle Triassic fish faunas in both the Western and Eastern Tethys, apart from the Besano Formation in the Monte San Giorgio area [35,36,37]. The recovery from bed 96 of part of the body of a shark, with a patch of skin packed with placoid scales, some calcified cartilage, and a dorsal fin spine, is to be considered outstanding, since in other Ladinian fish faunas from the Alps, no remains at all are recorded (as for the Chinese Xingyi Fauna), apart from small isolated teeth usually found in the leftovers of samples dissolved to search for conodonts. In this assemblage, a few teeth, probably belonging to Palaeobates angusissimus (Agassiz, 1834) [38,39], have also been found, so we can assume that the body fragment (still under preparation) could belong to this same species. The presence of such a durophagous shark in the Pelsa/Vazzoler lagoon is probably related to the richness in small invertebrates that thrived in the shallow water surrounding the depositional basin itself [27], while strictly durophagous actinopterygians have not yet been recorded.

Coelacanth remains are represented by two small articulated specimens and several isolated scales, some of them quite large, being over 2 cm in diameter.

A few teeth and scales have also been recovered from beds 98, 97, and 93, which yielded the silicified fauna after limestone dissolution by acetic acid. Again, actinopterygians, coelacanths, and sharks are also represented in these beds, which are herein considered as having originated from mass transportation from the shallow waters of the margin of the lagoon.

Thus, the Pelsa/Vazzoler Fossil-Lagerstätte is proving to be one of the major Triassic deposits because it is rich in both marine vertebrates and invertebrates, as well as in terrestrial organisms (plants and insects), whereas most of the other marine Triassic Fossil-Lagerstätten mainly yield only one such group of organisms.

Herein, Thoracopterus wushaensis will be described for its taxonomic, stratigraphical, and paleobiogeographical significance. Photographs were taken in the Photographic Laboratory of the Department of Geosciences of the University of Padua, using a Sony (Tokyo, Japan) a7R III camera and Sony (Tokyo, Japan) FE 90 mm f/2.8 macro and Canon (Tokyo, Japan) MPE 65 mm f/2.8 lenses.

3. Paleontological Description

Actinopterygii Cope, 1871

Peltopleuriformes Gardiner, 1967

Thoracopteridae Griffith, 1977

Emended diagnosis. (* for apomorphic character inside the family): Small to medium-sized fishes. Skull roof bones randomly fused, usually with broad paired frontals, but large parieto-dermopterotic sometimes fused together; no distinct parietals. Nasals separated by a large convex rostral. Supraorbital single or few in number. One suborbital. Opercular series with large opercle, smaller subopercle, several branchiostegal rays*, lateral and median gulars. Vertical, very narrow preopercle abutting the rear edge of the maxilla with the infraorbital process*. Maxilla extending beyond posterior border of orbit and with large postero-ventral plate. Mandible with very low coronoid process. Spiracular bones present. Large otoliths present*. Sensory canal system reduced. Posterior preural neural spines expanded substantially longitudinally*. Body totally or partially scaled, or naked. Tail semiheterocercal, with more than ten epaxial rays*. Caudal fin deeply forked, with lower lobe much larger than upper**. Pectoral and pelvic fins very large and elongated (over 30% of s.l.)*. Anal and dorsal fins well posteriorly inserted; lepidotrichia reduced in length in the dorsal fin*. Anal fin sexually dimorphic.

Remarks. Griffith [3] considered Thoracopterus as belonging to the Luganoiformes (together with his “Habroichthys” gregarius Griffith, 1977 [3]). Lehman [4] previously noted that Thoracopterus has nothing to do with Luganoia and Besania and set Thoracopterus close to Perleidiformes. Tintori and Sassi [5] considered Thoracopterus related to Peltopleurus (thus belonging to Peltopleuriformes) rather than to Perleidus (Perleidiformes). Since then, these flying fishes have been included in this order.

Shen and Arratia [40] erected the superfamily Thoracopteridea, as well as a few new generic taxa for the Thoracopteridae, claiming that Thoracopterus is not “monophyletic,” but without giving a clear explanation. The author does not consider their analysis to be valid, or their taxonomic conclusion, because the Thoracopteridea diagnosis shows several inconsistencies and mistakes, as shown below (in italics) where the characteristics they used in the superfamily diagnosis are discussed, and the matrix by Shen and Arratia [40] used for the phylogenetic analysis is not at all precise with regard to many characteristics (see below).

“Small (c. 60 mm) to medium-sized (c. 190 mm) fishes”. Most Triassic actinopterygian fishes are within this range, apart from the large predatory taxa such as Saurichthys and Birgeria. It may be better to retain it inside peltopleuriforms so that this size range may have real meaning, although Wushaichthys is out of this range, as in Shen and Arratia [40], it is said to be a maximum of 57 mm in total length, with most specimens being much shorter (pers. obs.). Additionally, the length of Peripeltopleurus is, on average, below the lower boundary of 60 mm [41], so the given range seems to be correct for Thoracopteridae but not for the other taxa considered in the new superfamily.

“Frontal bones are unfused; frontal bones are broad and laterally expanded at orbital level*”. Frontals are unfused in most actinopterygian fishes, as fusion among skull roof bones is quite a rare event, and usually it is erratic even inside the same species, as for Thoracopterus martinisi, where the skull roof bone can be totally fused and there is also a narrowing at the orbital level [5], or for T. “niederristi” from Polzberg [3] (pers. obs. 2025), which is actually a new species, different from the one from Raibl (Tintori and Conedera, submitted). In Shen and Arratia [40] (Figure 9A,B), the frontals in Wushaichthys appear as if they can fuse to each other. The same is observed in Peripeltopleurus [40] (Figures 154B and 155A). Similar variability in fusion of the skull roof bones is shown in several peltopleuriforms, such as Habroichthys [10,33], P. tyrannos [42], and P. nitidus [43], or in advanced neopterygians, such as Prohalecites [44]. Thus, this appears to be a general characteristic of actinopterygians, but not for this superfamily or for Peltopleuriformes, where frontal fusion is quite common.

“Parietal bone and dermopterotic are fused into a large parieto-dermopterotic”. See above.

“No postrostral bone in skull roof. Nasal bones are separated medially by a large rostral bone”. The postrostral is present in very few basal actinopterygians, and apparently only in Colobodus among “subholosteans” [45,46]. No other peltopleuriform or perleidiform has this bone, and in all peltopleuriforms, the large rostral totally separates the nasal bones [12,47,48,49], while in perleidiforms, the nasals are usually in contact shortly after the rostral [11,49,50]. It must also be noted that Shen and Arratia [40] accepted the restorations of T. “niederristi” and T. telleri made by Griffith [3] and Lehman [4], where the nasals meet in front of the frontals, clearly in contrast with this diagnostic character. However, Tintori and Sassi [5] reported that the restorations were wrong (see also Tintori and Conedera, submitted). However, this is a general characteristic of Peltopleuriformes.

“Posttemporal contacts extrascapular anterolaterally and separates this bone from contacting its counterpart*”. This is not the case for T. magnificus and T. martinisi [5], whereas in T. telleri, the area is not known. In T. “niederristi” from Polzberg, both Griffith [3] and Lehman [4] consider the two bones medial to the lateral extrascapular as “median extrascapular”. In Wushaichthys, Shen and Arratia [40] (Figure 10A), the posttemporal is just posterior to the extrascapular. Thus, this characteristic, even if considered synapomorphic for the superfamily, does not appear to be present in almost all the taxa.

“Spiracular bone(s) present”. This is a very general observation in most actinopterygians.

“Skull roof sensory canal system short, simple and discontinuous, with a few small sensory pores opening directly on the surface of the canal”. This is also the case in many other peltopleuriforms and perleidiforms, for instance, Peltopleurus and Habroichthys [10,33,43,48,51].

“Elongate maxilla extending beyond the posterior margin of the orbit”. All other Peltopleuriformes (and Perleidiformes) show this characteristic. This is the standard state in all non-neopterygian actinopterygians, and also some advanced Triassic neopterygians, such as most Pholidophoridae [14,17] and the Panxianichthyformes halecomorphs show this character, so the meaning of this characteristic for the superfamily can be questioned; thus, it is not clear what the significance of this characteristic is for the superfamily.

“Enlarged antorbital, nearly equal to or deeper than the nasal bone”. Not in all species.

“One or a few supraorbital bones. Few to several suborbitals present”. These are common states for most actinopterygian fishes. Regarding peltopleuriforms, in all Thoracopterus species, these bones are not usually recorded, or for suborbitals, there is a single large bone [5,15], as for Peltopleurus specimens [15,48]!

“Lower jaw with low coronoid process or process absent”. Again, this is a general non-neopterygian characteristic.

“Surangular bone commonly missing”. Alternatively, you cannot clearly see this region in such small fishes with no coronoid process.

“Opercle more than 2.5 times as large as the subopercle”. Many other peltopleuriforms, such as Habroichthys, a few Peltopleurus, and Nannolepis, show a very small subopercle [3,10,33].

“Preopercle narrow and a deep rectangular bone, vertically oriented *. Preopercle suturing with rear edge of the maxilla throughout the maxillary process*. Preopercular sensory canal simple, with one short branch extending into the maxillary process. Suspensorium vertically oriented”. These are common characteristics in almost all Peltopleuriformes, with only Habroichthys being somewhat different in having a very short suture with the maxilla [33]; thus, these characteristics are Peltopleuriformes synapomorphies.

“Interopercle absent”. Again, this is true for all non-neopterygian actinopterygians and thus also for all Peltopleuriformes and Perleidiformes.

“Median and lateral gulars present in most specimens”. A diagnostic characteristic should be present in (almost) all the considered taxa, not specimens (see below for several other characteristics where Shen and Arratia [40] used “most genera” or “generally”). Using “specimens” makes things unclear. Furthermore, gulars are present in almost all non-neopterygian actinopterygians, although they are usually not recorded in many miniature peltopleuriforms but are present in Perleidus altolepis (Deecke, 1889) [49,52] and Colobodus [46].

“Enlarged pectoral fins in most genera”. The term “enlarged” can be interpreted in many ways; for instance, in Wushaichthys exquisitus, they are in the same proportion to standard length as in Peltopleurus nuptialis Lombardo, 1999 [40,48], as in almost all other Peltopleuriformes. However, they are much shorter than in “true” flying fishes (Thoracopterus spp. or Thoracopteridae), where pectorals should be at least around 30% of the fish’s standard length [9,15].

“Pelvic fin well developed in most genera”. See above.

“Fringing fulcra on leading margin of paired fins generally absent. Fringing fulcra on leading margin of the dorsal and anal fins absent”. This characteristic appears to be very widespread among peltopleuriforms. Peltopleurus shows only very small fringing fulcra on both margins of the caudal fin [47,48], as for W. exquisitus [40]. Altisolepis lacks fringing fulcra in all fins [12], as is also the case for Habroichthys [10,33] and Nannolepis [3] (pers. obs., 2025). Thus, again, this is a Peltopleuriformes characteristic and is also shown in a few Perleidiformes, such as Dipteronotus [47].

“Dorsal and anal fins generally posteriorly inserted, closer to caudal fin than the midpoint of body length”. As in many other Triassic fishes, and also in most Peltopleuriformes.

“Dorsal and anal fin rays reduced in length”. As in many other Triassic fishes.

“Caudal fin forked, with its hypaxial lobe longer than the epaxial lobe*. Hypaxial lobe of caudal fin broader than epaxial lobe”. The actual ratio between the two lobes is quite different in Thoracopterus with respect to Wushaichthys or Peripeltopleurus.

“Scales (when present) of the main horizontal row, along the lateral line, much deeper than scales above and below the lateral-line scales”. All scaled “subholosteans” (Peltopleuriformes, Perleidiformes, Pholidopleuriformes) share this characteristic. A few Middle Triassic neopterygian fishes, such as Placopleurus, show a high flank scale row [53] (Conedera et al., submitted).

Genus Thoracopterus Bronn, 1858

Type species: Thoracopterus niederristi Bronn, 1858.

Other species: T. telleri (Abel, 1906), T. magnificus Tintori and Sassi, 1987, T. martinisi Tintori and Sassi, 1992, T. wushaensis Tintori et al., 2012, and T. sp. n. Tintori and Conedera, submitted.

Diagnosis: As for the family because it includes only one genus.

Age and geographical distributions: Late Ladinian (Middle Triassic) of the Western Tethys (Dolomites, Belluno Province, northeastern Italy) and Eastern Tethys (Xingyi area, Guizhou Province, southwestern China) to the Late Triassic of the Western Tethys (early Carnian of Austria and early Carnian to Norian of northern Italy).

Remarks. Since the original description of T. wushaensis by Tintori et al. [15], a few authors have dealt with this species, the other Triassic flying fishes, and the related peltopleuriforms, starting from Xu et al. [54] with the description of other specimens, not collected in situ, under the name Potanichthys xingyiensis, down to Shen and Arratia [40]. However, as will be discussed in the following paragraphs, the anatomical descriptions, the characteristics identified, the comparison with other taxa, and the related phylogenetic analyses are herein considered too scant to support their conclusions from a nomenclatural point of view. Several of these problematic aspects have already been pointed out in Tintori [55]. For their analysis, Xu et al. [56] used 23 taxa with 80 characteristics, while Shen and Arratia [40] used 137 characteristics from 54 taxa. A random check was performed on Xu et al.’s [56] matrix on a few taxa the author is familiar with (Habroichthys, Venusichthys, Peltopleurus-Peripeltopleurus-Wushaichthys), and the results are shown further below. The high number of wrongly coded characters, other than the several missing ones, together with other problematic assumptions, leads the author to discard this analysis and those following as an upgrade of the original one. A check of the most recent analysis [40] also unveils problems in the coding of the Thoracopteridae species and, for T. niederristi (the new species from Polzberg, so the type species is not considered), shows 40 unknown characteristics, which are reduced to 39 after checking and changing 27 wrongly coded ones; for T. wushaensis, 23 were unknown, now 26 with 17 wrongly coded ones; for T. telleri, 64 were unknown, now 57 with 29 wrongly coded ones; for T. magnificus, 53 were unknown, now 49 with 9 wrongly coded ones; and for T. martinisi, 49 were unknown, now 44 with 17 wrongly coded ones. All this makes the statistical analysis very poor and not suitable to be trusted.

The geographic and stratigraphic positions of the Chinese material described by Xu et al. [54], Xu et al. [56], Xu and Zhao [53,57], and Shen and Arratia [40] are imprecise, as Tintori [55] already pointed out. None of those papers provide a studied section, record, or number of beds yielding the studied specimens. Moreover, the photographs of the outcrops only show some bedded dark limestone, sometimes moss-covered [40,57], that could belong to any of the several bedded limestone units in southwestern China (or elsewhere), or they show one of the authors [57] without a hammer beside a pile of slabs, a typical remnant of excavations by locals. We must also keep in mind that the only scientific excavations in the Xingyi area have been those in Nimaigu (Wusha District, Xingyi City) and in Dingxiao, carried out by the international team led by PKU paleontologists, from which all the specimens are recorded with the bed number, such as the case of T. wushaensis [14,15]. The Dingxiao locality is very close to the site from which the first fossils of the Xingyi Fauna were described [19], while the large Nimaigu excavation is about 50 km to the west. This detailed stratigraphic position is very important in highly variable groups such as Peltopleuriformes, as different species can very often be recorded in a very short stratigraphical interval of only a few beds, as the author could observe for Peltopleurus in the Kalkschieferzone from Monte San Giorgio, Habroichthys [33], or also for “Wushaichthys” in the Nimaigu excavation itself (pers. obs.). In the latter sequence, peltopleurids other than H. orientalis are quite common all along the fossiliferous level, which is subdivided into two major fossil assemblages, each one characterized by a very different array of taxa [14]. Although peltopleurids seem to be more common in the Upper Assemblage, a few specimens have also been found in bed 35, together with the very common H. orientalis, and below (pers. obs. during 2012–2013 fieldwork, when the author spent more than two months there in the bed-by-bed excavation). The type material of T. wushaensis is from this scientific excavation [15] and not “endemic to the village of Xiemi, Wusha district, City of Xingyi” [40], as Xu et al. [54] described Potanichthys xingyiensis. As already pointed out by Tintori [55], around Xiemi Village, no official excavations have been carried out, so the actual source of the Xu et al. [54] material is still in question. However, as P. xingyiensis is a junior synonym of T. wushaensis, as already stated in Tintori et al. [20] and Tintori [55] and accepted by Shen and Arratia [40], this species cannot be endemic to Xiemi, but at present is to be considered part of the Upper Assemblage of the Xingyi Fauna [15] and is probably known from a few different localities exploited by local fossil dealers, rather than from the Nimaigu type site. Shen and Arratia [40] used the generic name “Potanichthys” for the species Thoracopterus wushaensis. However, as “Potanichthys xingyiensis” is a synonym of T. wushaensis, the use of the generic name should also be avoided, but this is unimportant, as the author does not consider the analysis by Shen and Arratia [40] to be valid on the basis of several critical problems shown herein. It is also clear that the stratigraphy is not a strong argument for Shen and Arratia [40], as they wrote that Wushaichthys (which comes from the same stratigraphic unit as T. wushaensis, although possibly a few meters below in the late Ladinian) is one of the oldest representatives of Thoracopteridea, but Peripeltopleurus from the Besano Formation in Monte San Giorgio is from around the Anisian/Ladinian boundary, which is at least 2 Myr older! Thus, their chapter about “Evolutionary changes of fins” [40] (p. 22) is not so sound.

None of the authors involved in those papers [40,53,54,56,57] made a direct examination of European thoracopterid material held in Vienna, mostly housed at the Natural History Museum (the new species from Polzberg, as well as other specimens of T. telleri) or in the former Geological Survey, now GeoSphere (original material of the type species T. niederristi from Raibl/Cave del Predil, T. telleri holotype), in Bergamo and Milan (T. magnificus, original material and unpublished one), and in Udine (T. martinisi original material and unpublished one). Although T. telleri and T. “niederristi” from Polzberg have been redescribed by Griffith [3] and Lehman [4], their morphological descriptions were doubtful in several important aspects, as already pointed out by Tintori and Sassi [5]. This uncertainty is mostly due to the poor preservation of several specimens, especially the much rarer T. telleri [3,6] (pers. obs., 2013, 2025), in which the skeletal elements are usually not easily identifiable. Consequently, even Shen and Arratia [40] provided only very scant information and did not describe the holotype or any other specimen showing the skull. This explains why, in their analysis of 137 characteristics, a total of 86 are missing or wrongly coded. To further show how problematic the anatomical interpretation of this species is, Shen and Arratia [40] (Figure 6B) are clearly wrong in identifying as a “pelvic fin” (bottom left of the slab) the scattered hypaxial lobe of the caudal fin, while the other pelvic fin, probably the left if we consider the specimen as viewed from the ventral side, is partially visible just above the right one, thus in anatomical position, across the slab fracture (this is even more clearly visible in its counterpart NHMW 2007z0170/0206 (pers. obs., 2025)). The structures interpreted as “chordacentra” [40] (Figure 6C) are also very doubtful, as in no other specimen are they visible when the vertebral column is preserved, and in this same specimen, scarce traces of the vertebral column are present only between the pectorals, but no chordacentra can be recorded. However, in their analysis, Shen and Arratia [40] considered vertebral centra (autocentra or chordacentra, characteristic 95) as missing in T. telleri. Griffith [3] (Figure 7C) interpreted them as a “vestigial scale row” and, mostly owing to their shape and the presence of ganoine tubercles (pers. obs. on the counterpart, 2013, 2025), the author agrees with this interpretation. On the other hand, Shen and Arratia [40] (p. 9) themselves underlined that “the anatomical information on thoracopterids is incomplete due to poor preservation of the specimens”; thus, working from pictures does not seem to be the best way to improve their knowledge.

We must also keep in mind that both Griffith [3] and Lehman [4] did not check the topotypical material of T. niederristi, but they just trusted Abel [6]. In fact, it seems that there are a few differences between the Raibl topotypical T. niederristi material and the Polzberg one described by Griffith [3], Lehman [4], and Shen and Arratia [40], so the author assumes that they belong to two different species (pers. obs., 2025), leading to the erection of a new species (Tintori and Conedera, submitted). Thus, the T. niederristi in the recent analysis [40,53,54,56] is actually a different species (Tintori and Conedera, submitted) and not the type species of the genus Thoracopterus. This is also supported by the fact that no species are common between the two fish assemblages (Raibl and Polzberg), which are actually quite different from each other (pers. obs., 2013, 2025) and are also of different ages (Tintori and Conedera, submitted). Shen and Arratia [40] (Figure 6A) show only a picture of a Polzberg specimen of T. “niederristi”, but their “dorsal fin” is actually the distal part of the right pectoral fin (pers. obs., 2025). Furthermore, all the references in the figures are only related to the fins without any details on the other characteristics used in the analysis, casting further doubt on their coding.

The Italian Norian species, despite being described more recently, lack the detailed description needed to code the 137 characters used by Shen and Arratia [40], and they probably need to be redescribed using sound phylogenetic analysis, which is not the aim of the present paper. We must also keep in mind that most other peltopleurids are miniature fishes, 20–30 mm long, totally flattened, so it is almost impossible to check all the characteristics that are mostly related to neopterygians or teleosts. The author must finally point out that the major shared derived characteristic of the “subholosteans” (mainly Peltopleuriformes and Perleidiformes) is not present in all the considered analyses, i.e., the hemiheterocercal caudal fin with epaxial rays [49], which is not recorded in any true neopterygian. Xu and Zhao [43] describing P. nitidus, used “almost homocercal”, as did Shen and Arratia [40], which is not clear in its meaning, other than that the tails are “almost symmetrical”, although the terminology heterocercal, abbreviated heterocercal = hemihomocercal, homocercal, or hemiheterocercal, is based on the relationships between the internal structure and the rays. Perhaps “almost homocercal” could be used for advanced non-teleost neopterygians, such as Marcopoloichthys or Placopleurus, where the dorsal body lobe is extremely reduced to just a couple of scales (urodermals) [21] (Conedera et al., submitted). Without considering this synapomorphic character shared by at least Peltopleuriformes, Perleidiformes, and Pholidopleuriformes, the basis of the analysis appears to be not so sound. Additionally, the enlarged neural spines in the caudal region of T. magnificus, T. wushaensis, and T. martinisi, which are considered a synapomorphic characteristic of all thoracopterids, being strongly related to the flying capability [5,15], are not considered at all.

For Norian Thoracopterus species, the age is wrongly stated by Shen and Arratia [40], as T. magnificus is said to be of “Middle Triassic” age, while it is from Norian (Late Triassic) sites of the Bergamo province, as well as from Seefeld (Austria) and Giffoni (Naples), not just from Endenna-Zogno. T. martinisi is from Norian sites too, but possibly somewhat older than those yielding T. magnificus, not simply “Late Triassic”, as stated by Shen and Arratia [40] (the Late Triassic is 35 Myr long!), and comprises also the Carnian stage, which T. niederristi, T. telleri, and the new species from Polzberg are from. Regarding the size of the two Norian species, Shen and Arratia [40] wrote that they are about 80 mm long. However, T. magnificus reaches at least 135 mm [9], and T. martinisi at least 200 mm (pers. obs. on undescribed material stored in Udine Museum).

For Thoracopterus comparison, Lehman [4] used the Early Triassic “Perleidus” species from Madagascar, which has since been proven not to be a Perleidiformes [11,49,58,59]. Unfortunately, most other authors, and among them Xu et al. [54], Xu et al. [56], Xu and Zhao [53], and Shen and Arratia [40], followed Lehman [4], not caring about the recent work on the type species of Perleidus [49,59] or new species from southwestern China [11]. Again, the presence of epaxial rays that differentiates the “subholosteans” from neopterygians is not considered (see, for instance, Xu et al. [60]).

Venusichthys (Xu and Zhao, 2016) is considered a peltopleuriform based mainly on the very high flank scales, although Shen and Arratia [40] put it outside Peltopleuriformes but still very close to Habroichthys, always considered a peltopleuriform, although in a family of its own, Habroichthyidae, as in Gardiner [61] (see also Conedera et al. [33]). After checking material from both ends of the Tethys (the Luoping Fauna and different new Alpine sites), the author considers Venusichthys to be a junior synonym of Placopleurus Brough, 1939 [62], due to their common unique scale ornamentation—very thin longitudinal striae on the anterior region, the one covered when in anatomical position that was not cited by Xu and Zhao [53] but is present in Chinese specimens from the Luoping Fauna, as well as the peculiar preopercular and lower jaw structures (pers. obs.). Xu and Zhao [53] wrote that their new genus is different from “the type species of Placopleurus (P. primus) from the late Ladinian (latest Middle Triassic) of Besano, Italy, as well as other known peltopleurids”. In fact, Brough [62], when erecting Placopleurus, used P. primus as the type species based on a single very poorly preserved specimen, but ascribed a previously described species (Pholidophorus besanensis Bassani, 1886 [63]) to his new genus based on several specimens, which were not well preserved. The presence of the characteristic scale ornamentation is described for P. besanensis, but it is also clearly visible on the imprint of some of the scales of the highly disarticulated holotype (and only specimen) of P. primus (C. Lombardo, pers. com., 2004, Conedera et al., submitted). Furthermore, there is no typical peltopleuriform sexual dimorphism in Placopleurus, as stated by Xu and Zhao [53] (pers. obs. on both European and Chinese specimens), as there is no modification in the anal fin of the supposed males, and the caudal fin does not show any epaxial rays. In their phylogenetic analysis, for “Venusichthys”, Xu and Zhao [53] coded 80 characteristics, 22 of which were unknown or not available, and 12 were wrongly coded (1, 2, 31, 39, 44, 48, 52, 53, 54, 62, 65, and 70)! Additionally, the age of the Besano Formation is not late Ladinian (as stated by Xu and Zhao [53]) but around the Anisian/Ladinian boundary. Thus, “Venusichthys” being Placopleurus is not a peltopleuriform at all, as it lacks, first of all, the hemiheterocercal tail with epaxial rays, and it is instead a non-teleost neopterygian (Conedera et al., submitted).

Regarding Habroichthys, another iconic miniature peltopleurid fish from the Middle Triassic, the published analyses are, again, quite rough and unreliable. In the starting analysis by Xu and Zhao [53], out of 80 characteristics, for Habroichthys, 29 were unknown or NA, and 20 were clearly wrongly coded (1, 2, 18, 23, 30, 31, 33, 36, 44, 46, 48, 49, 50, 51, 52, 53, 56, 65, 70, 79). This means that only 31 out of 80 characteristics (38%) have been correctly coded, most of them totally uninformative from a phylogenetic point of view, and 25% were wrongly coded! Shen and Arratia [40] upgraded the analysis using 137 characters (please, again, keep in mind that Habroichthys specimens are mostly less than 30 mm long and totally flattened, and the authors did not check any new specimens), but this did not change Habroichthys’ position. It must also be pointed out that Xu and Zhao [53] cited H. orientalis (Su, 1959) as Peltopleurus, not considering the main Habroichthys derived characteristic, i.e., the last flank scale being enlarged and semicircular [62]. Lin et al. [10] and Tintori et al. [64] moved this species to Habroichthys because of this generic apomorphic character, which is not shown in any other peltopleuriforms but was not considered by Xu and Zhao [53]. A new array of Habroichthys specimens from various sites and ages is now under study [33] (Tintori et al., submitted), but the average very small size and preservation quality will not vastly improve skull anatomy knowledge.

Other problems arise from the group of genera Peltopleurus/Peripeltopleurus/Wushaichthys that can be considered very close to each other, as can be seen from the diagnoses [41,56] that are totally uninformative and do not explain the differences among the three genera. As for Peltopleurus, in their analysis, Xu and Zhao [53] did not consider the type species, Peltopleurus splendens Kner, 1866 [16], from the early Carnian of Raibl/Cave del Predil (Italy) but referred to published material of P. rugosus Bürgin, 1992 [41], from the Besano Formation (Monte San Giorgio), basing their observation on photographs or on the Chinese species P. nitidus Xu and Ma, 2016, and P. tyrannos Xu et al., 2018 [40]. However, if these two latter species have been ascribed to Peltopleurus, it is not clear why a new genus, Wushaichthys, has been erected for the species W. exquisitus. P. tyrannos, for instance, appears to be much closer to Perleidiformes than to peltopleurid in terms of scale pattern, preopercular shape, dentition, etc. However, regarding Wushaichthys, it has been coded in Xu and Zhao [53] with 22 unknown characteristics and 13 wrong ones (1, 15, 18, 22, 31, 44, 46, 48, 49, 54, 57,75, 79), while for Peripeltopleurus, there are 22 unknown characteristics (but two are different from Wushaichthys) and 11 wrong ones (1, 9, 15, 18, 31, 44, 46, 48, 55, 57, 79), with only 56% and 58% of the characteristics being correctly coded. In fact, the characteristics that separate Wushaichthys and Peripeltopleurus (clavicle and antorbital) are not easily visible in such small specimens, as has already been discussed herein. Finally, Shen and Arratia [40] wrote that the Wushaichthys specimens were from “near the village of Nimaigu”, although in Xu et al. [56], it is reported as simply “Wusha”. What is certain is that the Wushaichthys material of the type series and the new one described by Shen and Arratia [40] are not from the official PKU excavations in Nimaigu. Thus, again, the real origin of the specimens is not clear, and, furthermore, it is not known if they are from a single site and/or from one bed or several different ones.

In conclusion, the author does not consider these analyses valid, following the above discussion, and the author considers Thoracopterus as a unique genus of the Thoracopteridae, comprising at least six species (T. “niederristi” from Polzberg is now considered a different species by Tintori and Conedera, submitted), considering four different genera for six species.

Thoracopterus wushaensis Tintori et al., 2012

2012 Thoracopterus wushaensis Tintori, Sun, Lombardo, Jiang, Ji, and Motani: p. 42, Figures 2–10.

2013 Potanichthys xingyiensis Xu, Zhao, Gao, and Wu: p. 2, Figure 1.

2014 Thoracopterus wushaensis Tintori et al.: p. 401, Figure 8

2015b Thoracopterus wushaensis Tintori: p. 1.

2024 Potanichthys wushaensis comb. nov.: Shen and Arratia: p. 9, Figure 6D

Material. A single specimen in part and counterpart, MGP-PD 33418a, b, stored at the Museo della Natura e dell’Uomo (Geo-Paleontological collections) of the University of Padua (Figure 3).

Stratigraphic and geographical distributions. The type locality is the scientific excavation near the village of Nimaigu (Wusha District, Xingyi City, Guizhou Province, China), Upper Vertebrate Horizon, lower part of the Zhuganpo Member of the Falang Formation, late Ladinian (Middle Triassic), although it is probably present in a larger area surrounding Wusha. The Pelsa/Vazzoler Fossil-Lagerstätte, Site A, bed 96 (“fish bed”), nearby Casera Pelsa (Taibon Agordino municipality, Agordino Dolomites, Belluno, Italy), Punta Santner Member of the Sciliar Formation (Longobardian, late Ladinian, Middle Triassic).

Description. The new Italian specimen is an incomplete one (part and partial counterpart) showing most of the body but lacking the skull and the anterior part of the vertebral column. The estimated standard length is around 110 mm, and the total length around 130 mm, well inside the range for the type series s.l. (87 mm to 117 mm).

The vertebral column. The vertebral column of the specimen is visible, apart from the most anterior and the most posterior parts (Figure 3B and Figure 4A). In the thoracic region, the elongated paired neural arches show an anterior triangular process at the distal third. Supraneurals are present in the most anterior part (as in the type material [15], probably down to halfway between the skull and the beginning of the dorsal fin, which is quite posteriorly inserted. Supraneurals are thin and much longer than the corresponding neural arches. Here, the haemal arches are strongly built and appear to make an almost continuous canal along the ventral side of the notochord to the first radial of the anal fin, which is very elongate (see below). Small bony splints are also present, and they have been interpreted as short ribs in the type material [15]. In the region between the first anal radial and the beginning of the dorsal fin endoskeleton, the paired neural arches are shorter, and the haemal arches are smaller than the anteriormost ones, and still without a median spine that starts to appear in correspondence with the first dorsal fin radial. In this region, the vertebral column is clearly diplospondilous. Median neural spines are present from below the beginning of the dorsal fin [15], although, unfortunately, the posterior part of the caudal region is not well preserved owing to the detachment of the caudal fin from the supporting endoskeleton.

Pectoral fins. Although not well preserved, at least 12 rays are visible, the anteriormost of which are very short, like in the type material. The surface of the proximal elements is ornamented with longitudinal ridges of ganoine. The preserved part, probably lacking only the very proximal end, is 55 mm long (Figure 3).

Pelvic fins. The pelvic fins are much shorter than the pectorals, reaching 24–25 mm in length. Lepidotrichia, around 10 in number, are characterized by a very long proximal element, followed by a segmented region with two branches, the last one very distal (Figure 3B and Figure 4B). The first ray seems to be shorter than the maximum length of the fin and unbranched. Thin ganoine ornaments are present at least on the proximal elements. Very large triangular plates make up the endoskeleton of the fins.

Anal fin. The well-preserved anal fin shows the typical peltopleuriforms “male” structure [48] (Figure 9A), wrongly interpreted as “female” in the paratype GMPKU-P-3071 by Tintori et al. [15]. The endoskeleton is made of about 10 radials, probably subdivided into elongate axonosts and very small, rounded baseosts. The axonosts are very large and elongate, especially the first one, its proximal tip being close to the haemal arches. They are very forward inclined, as is also seen in the type material from China [15]. Baseosts are round, their diameter corresponding to the distal width of the corresponding axonost. Externally, two large oval scales cover each side of the proximal region of the lepidotrichia of the anterior region. These scales are fully ornamented with low ganoin tubercles. As is typical in “male” peltopleuriforms, the external part of the fin is subdivided into three regions (Figure 3B and Figure 4D). The anterior part is made up of nine segmented rays that become somewhat thinner backwards. They are somewhat shorter than the median region rays, and they branch only very distally, at least twice, all elements remaining in contact with each other. The middle region, clearly spaced from the anterior one, is made of two sets of lepidotrichia that show, as common characteristics, a few stout proximal segments followed by a section made up of very short segments (the “hinge area” in Lombardo [48]). At least seven unsegmented rays emerge from the hinge area, making up the first group of the middle region. They are quite thin and branch distally, possibly twice, the distalmost elements bending forward. The second group of rays in this region shows the posteriorly directed hooks. Seven to eight segmented rays have the distal element modified as a large, stout hook, making an almost vertical posterior edge. The preservation does not allow precise restoration of the posterior region, but a few proximal segments (possibly three) are followed by short lepidotrichia, of which only scattered segments are visible. No fringing fulcra are present.

Dorsal fin. The dorsal fin is inserted well posterior to the anal fin, the first radial being around the vertebral column position of the last one of the anal fin (Figure 3B and Figure 4A). At least 10 radials are present, the first one being much larger than the following ones and with a bifurcated proximal region, in the same way as in the type series [15]. As in the latter, this large radial appears to be just in front of the lepidotrichia, without giving support to any of the long rays. Owing to the twist of the posterior caudal region, rays are packed together, and no detailed description is possible.

Caudal fin. Owing to the twist of the vertebral column in the tail region, only the ventral lobe and part of the dorsal one are well preserved, the dorsalmost part of the caudal fin being preserved as scattered rays mixed with disarticulated scales from the left side of the caudal scale field. The fin is deeply forked, the ventral lobe rays being much stouter and packed than those in the dorsal one. The ventral margin is made of a few basal fulcra (possibly five) followed by four short, segmented elements. The leading ray is the longest, and it is followed by another four somewhat large rays of more or less the same length, all of them branching, possibly twice, only very distally. After these five rays, the other five in the ventral lobe become quickly shorter, and their branching starts very proximally after a very enlarged proximal triangular element. Thus, the ventral lobe is made up of 10 strictly packed rays with long segments, while the posterior margin is a fringe of very thin elements owing to strong distal branching. No fringing fulcra are visible. The median part of the dorsal lobe shows two rays similar to the last two in the ventral lobe, but these dorsal rays appear somewhat separated from each other. As said before, only scattered rays from the remaining part of the dorsal lobe are preserved along the ural region of the vertebral column.

Scales. The body was almost totally naked, and, as this specimen lacks the anterior region, it is not possible to check the presence of the antero-dorsal scale field described in the Chinese specimens [15]. Only the caudal scale field is preserved ventrally to the vertebral column (Figure 3B and Figure 4C), as in the Chinese type series. The only difference appears to be in the number of rows, here 11. The first (anteriormost) two rows are much shorter, and the scales are smaller (8–9 scales per row), while the following four are the most important, with more than 10 large scales. The five posterior rows decrease in both the number of scales and the size of single scales. No peg and socket articulation is visible, and the scales appear to be just slightly superimposed on each other. The external surface is covered by a smooth layer of ganoine.

4. Discussion

The only difference from the type material of T. wushaensis is the number of scale rows in the caudal region (11 vs. 4 in the type material). However, if we consider the average preservation of that region, it seems believable to suppose that, in the Chinese specimens, we can observe mostly only the main four rows that are present in the new specimen, while the smaller anterior and posterior ones could be scattered owing to the lack of a strong articulation. Furthermore, in all specimens where this region is somehow preserved, the scale rows never appear to encircle the fish body, as shown by Xu et al. [54].

The anal fin sexual dimorphism shown by T. wushaensis is very similar to that of Peltopleurus spp. [48,65], strengthening the Peltopleuriformes relationships. It is possible that some differences also could be related to the different average sizes of most Peltopleurus species compared to Thoracopterus. Furthermore, the large size and the presence of several large bony hooklets on the posterior edge confirm the total unlikelihood, as already suggested [55], of the use of a similar fin as a gonopodium for internal fertilization [43,56].

5. Conclusions

The discovery of T. wushaensis at the western end of the Tethys further strengthens the relationships among marine vertebrate faunas across the Tethys itself (Figure 5). Along with a number of marine reptile and fish genera, a species is now described from both China and the Alps, and the author must highlight that this species is, at present, considered the oldest of the known flying fish Thoracopterus, which means that migration across the Tethys was probably not related to a single event [20], but that it was “always” possible for small fishes or large pelagic marine vertebrates to move east to west near the Equator and the Cimmerian continental masses, or west to east along a longer northern route, following the major oceanic currents (Tintori et al., submitted). However, fast swimming fishes, such as the flying ones, or the large palaeonisciform Guizhouniscus microlepidota Liu, 2003 [66,67], also from the Xingyi Fauna, could have been able to travel long distances without the current’s help. This is also the case with several modern flying fishes, such as the “monoplane” Exocoetus volitans [68] or the “biplane” Cheilopogon heterurus [69], which are found all around the world in tropical and subtropical seas. This is also true for marine reptiles [18], as oceanic ichthyosaurs, thalattosaurs, and pistosaurs show a close paleobiogeographic affinity to North America. Finally, it should be noted that benthic invertebrates also experienced a significant bloom at this same time, as evidenced by the very rich silicified fauna found in the beds around the fish layer yielding T. wushaensis from the Pelsa/Vazzoler Fossil-Lagerstätte [27]. This indicates that in the late Ladinian, we are likely observing a major marine bioevent during the post-Permian/Triassic crisis recovery.