Rivista Italiana di Paleontologia e Stratigrafia (Research in Paleontology and Stratigraphy) vol. 128(1): 23-42. March 2022 THE PRESENCE OF AN ORBITOANTORBITAL FENESTRA: FURTHER EVIDENCE OF THE ANUROGNATHID PECULIARITY WITHIN THE PTEROSAURIA FABIO MARCO DALLA VECCHIA Institut Català de Paleontologia (ICP), Edifici ICP, Campus de la Universitat Autònoma de Barcelona, E-08193 Cerdanyola del Vallès, Spain. E-mail: fabio.dallavecchia@icp.cat Associate Editor: Silvio Renesto. To cite this article: Dalla Vecchia F.M. (2022) - The presence of an orbitoantorbital fenestra: further evidence of the anurognathid peculiarity within the Pterosauria. Riv. It. Paleontol. Strat., 128(1): 23-42. Keywords: Reptilia; Flying reptiles; Anurognathidae; Batrachognathus volans; Skull; Anatomy. Abstract - The anurognathids are peculiar pterosaurs characterized by broad skulls with very short rostra and broadly arched jaws. The presence of distinct or confluent external naris and antorbital fenestra in these ptero- saurs has been debated in the last years. The relatively well-preserved specimens of Batrachognathus volans show that the antorbital fenestra was confluent with the orbit forming an enormous orbitoantorbital fenestra. This feature is evident also in Jeholopterus ningchengensis. The consequent modification of the matrices of two recently published phy- logenetic analyses about the in-group pterosaur relationships shows that the Anurognathidae are a derived clade of non-monofenestratan pterosaurs. Anurognathidae (including also ‘Dimorphodon’ weintraubi according to the definition by Hone 2020) are still a scarcely known clade because only a few specimens have been adequately described in the literature. Introduction Pterodactyliformes according to Andres et al. 2014), the narial and antorbital fenestrae are confluent Pterosaurs are extinct archosaur reptiles that forming a nasoantorbital fenestra. lived during the Mesozoic Era and were the first Anurognathidae is a peculiar pterosaur clade vertebrates to evolve the powered flight. The oldest that has been recently reviewed by Hone (2020) pterosaurs found to date are Late Triassic (Norian, and Wei et al. (2021). According to the definition ca. 220 million years ago) in age, while the last are of Anurognathidae by Hone (2020) (see below) and latest Maastrichtian (ca. 66 million years ago) in age the phylogenetic hypothesis by Wei et al. (2021), the (Dalla Vecchia 2013; Dalla Vecchia et al. 2013). The clade includes eight monospecific named genera non-monofenestratan pterosaurs had a moderately and no more than 15 reported specimens (Wei et al. elongated skull with five fenestrae (openings) 2021). Until 2002, only three anurognathid species, in laterodorsal view: narial, antorbital, orbital, each represented by a single and incomplete de- lower temporal and upper temporal fenestrae. In scribed specimen were known: Anurognathus ammo- monofenestratan pterosaurs (Darwinoptera + ni Döderlein, 1923, Batrachognathus volans Ryabinin, 1948 and Dendrorhynchoides curvidentatus (Ji & Ji, 1998). Later, Jeholopterus ningchengensis Wang, Zhou, Zhang & Xu, 2002; Luopterus mutoudengensis (Lü & Received: June 22, 2021; accepted: September 27, 2021 Hone, 2012); Vesperopterylus lamadongensis Lü, Meng, 24 Dalla Vecchia F.M. Wang, Liu, Shen & Zhang, 2018; and Sinomacrops et al. 2009, 2014, and 2017). They occur low in the bondei Wei, Pêgas, Shen, Guo, Ma, Sun & Zhou, pterosaur tree, although not in a basal position, in 2021 were added to the list. the phylogenetic hypotheses by Unwin (2003) and Unnamed and only preliminarily described Lü et al. (2010a, 2012), while they are closer to the anurognathids are reported from the Sinuiju Beds Pterodactyloidea according to other authors. They (Lower Cretaceous) of North Korea (Gao et al. are the sister group of the Pterodactyloidea in the 2009) and the Middle Jurassic of China (Jiang et al. phylogenetic hypotheses by Dalla Vecchia (2009a, b) 2015). Also a partial specimen from the Middle Ju- and Andres et al. (2010), but it must be considered rassic of Mongolia (Unwin & Bakhurina 2000) and that the wukongopterids were not included in these Mesadactylus ornithosphyos Jensen & Padian, 1989 from analyses because they were unknown before 2010. the Upper Jurassic of USA (Bennett 2007) have Anurognathidae are nested within the Monofene- been reported as potential anurognathids. However, strata as the sister group of the Pterodactyloidea in Hone (2020) considers Mesadactylus ornithosphyos to the more recent phylogenetic hypotheses by Andres be a pterodactyloid. (2012), Andres & Myers (2013), Andres et al. (2014) The phylogenetic analyses by Britt et al. and Wu et al. (2017), while they are the basalmost (2018), Dalla Vecchia (2019) and Wei et al. (2021) monofenestratan in that of Wei et al. (2021). Britt have recovered Dimorphodon weintraubi Clark, Hop- et al. (2018) and Dalla Vecchia (2019) were the first son, Hernández, Fastovsky & Montellano 1998 as to include ‘Dimorphodon’ weintraubi in their analyses; the sister taxon of all (Wei et al. 2021) or some (Britt they found the anurognathids inclusive of ‘Dimor- et al. 2018; Dalla Vecchia 2019) of the anurognathid phodon’ weintraubi to be the sister group of a trichot- taxa listed above. This supports the view by Britt et omy of Sordes pilosus, Rhamphorhynchinae + Scapho- al. (2018) that Dimorphodon weintraubi does not be- gnathus crassirostris, and Monofenestrata. According long to Dimorphodon and should be referred to a dis- to Vidović & Martill (2018), Anurognathidae are the tinct genus. According to the definition of Anurog- sister group to Sordes pilosus + Sericipterus wucaiwaensis nathidae by Hone (2020), ‘Dimorphodon’ weintraubi is within Scaphognathinae sensu Unwin (2003); ‘Di- an anurognathid in these phylogenetic hypotheses. morphodon’ weintraubi is not included in this analysis. All the anurognathid specimens are small- Batrachognathus volans is an anurognathid pter- sized (maximum 900 mm in wingspan; excluding osaur described by Ryabinin (1948) based on the the only specimen of ‘Dimorphodon’ weintraubi, which holotype specimen “No. 52 - 2” (Ryabinin 1948: is larger), preserved crushed on slabs, incomplete 87). According to Bakhurina & Unwin (1995: tab. and/or partly disarticulated. For these reasons, their 1), the species is represented by two specimens; ac- delicate skeletal elements are difficult to identify and cording to Unwin & Bakhurina (2000: 423), they subject to subjective interpretations. These speci- are the holotype PIN 52-2 and the undescribed mens span the Middle Jurassic (Callovian or possi- PIN 2585/4a (which was found much later than bly Aalenian) to the Early Cretaceous (Aptian) and the holotype and is preserved on the same slab sur- come from central-eastern Asia and Germany; the face as the holotype of the non-monofenestratan basal ‘Dimorphodon’ weintraubi is slightly older (Toar- pterosaur Sordes pilosus described by Sharov 1971). cian) and comes from Mexico. All the specimens Apparently, Unwin et al. (2000: tab. 2) reported but the two from Germany are preserved into rocks measurements from a third specimen of Batracho- of continental origin; all the continental anurog- gnathus volans specimen (PIN 13), which has never nathids come from lacustrine facies, excluding ‘Di- been figured, described or reported elsewhere with morphodon’ weintraubi. this inventory number; this specimen is actually The position of the anurognathids (sensu the holotype PIN 52-2 (see Bakhurina & Unwin Hone 2020) within the Pterosauria is dramatically 1995: 216). The two specimens referred to Ba- different in the various phylogenetic hypotheses. trachognathus volans come from the Upper Jurassic According to Wei et al. (2021) there are at least five (Oxfordian-Kimmeridgian) Karabastau Formation main positions recovered for the clade and they (Karabastau Svita; Unwin & Bakhurina 2000: 423) add a sixth one. In some phylogenetic hypotheses, of the Karatau Ridge, Kazakhstan. Specimen PIN the Anurognathidae are the most basal pterosaurs 52-2 includes the better preserved anurognathid (Kellner 2003; Lü & Ji 2006; Bennett 2007; Wang skull reported to date, although it is slightly disar- Orbitoantorbital fenestra in the anurognathids 25 ticulated and partially concealed by other skeletal no need of an additional definition like that given by Hone (2020: 1681): all taxa more closely related to Anurognathus than Dimorpho- elements. PIN 52-2 has been figured in Wellnhofer don, Pterodactylus or Scaphognathus. However, Unwin’s definition was (1991: 101), Bakhurina & Unwin (1995: fig. 8), Un- formulated when only three anurognathid species were known and win & Bakhurina 2000: fig. 21.3), and Hone (2020: implies that taxa having the typical anurognathid bauplan (see be- fig. 5). The skull of PIN 2585/4a is complete and low) would fall outside the Anurognathidae if they turned out to be more basal than Anurognathus ammoni and Batrachognathus volans. Un- articulated, although it is not well preserved as PIN der Unwin’s definition, ‘Dimorphodon’ weintraubi would not be consid- 52-2; it has been figured in Unwin (2006: fig. 5.3 ered an anurognathid. Similarly, sensu Unwin (2003), Vesperopterylus lower left corner). Despite their importance in the lamadongensis would not be considered an anurognathid in the strict understanding of the anurognathid skull morphol- consensus tree by Lü et al. (2018), and Dendrorhynchoides curvidentatus would not be considered an anurognathid in the strict consensus ogy, these specimens have never been described in tree obtained from the matrix of Dalla Vecchia (2019) modified ac- detail after the report of the holotype by Ryabinin cording to the results of this paper (see below). To avoid problems, (1948). the definition of Hone (2020) was used in this paper. Of course, one could retain Unwin’s definition of the Anurognathidae and use This paper deals with the reinterpretation of other names (e.g., Anurognathoidea or Anurognathiformes) for the the orbital, antorbital and narial fenestrae of the more inclusive clades. However, naming new clades is beyond the anurognathid pterosaurs and its implications on the scope of this paper. Furthermore, it is wiser to wait for a more relationships of the anurognathids. The reinterpre- exhaustive description of the available anurognathid specimens and an increase of the coded characters in matrices (hopefully leading tation is based mainly on observations of the skulls to more stable results) before introducing new clade names. of Batrachognathus volans. The data matrix of Dalla Vecchia (2019) has been modified according to the results of this paper and then used to perform a Institutional abbreviations: CAGS, Chinese Academy of parsimony-based phylogenetic analysis in PAUP 4.0b10 for Micro- Geological Sciences, Beijing, China; MCSNB, Museo Civico di Sci- soft Windows (Swofford 2002) under the same conditions specified enze Naturali di Bergamo, Italy; MPUM, Museo Paleontologico del by Dalla Vecchia (2019). In the character state codings, specimen Dipartimento di Scienze della Terra dell’Università di Milano, Italy; CAGS–Z070/CAGS–IG-02-81 (see below) has been considered as NJU, Nanjing University, Nanjing, China; and PIN, Palaeontological belonging to Jeholopterus ningchengensis, following Wei et al. (2021). Institute, Russian Academy of Sciences, Moscow, Russia. Only Dendrorhynchoides curvidentatus has been coded instead of Den- drorhynchoides spp., because Dendrorhynchoides mutoudengensis was re- ferred to the distinct genus Luopterus by Hone (2020). Character 17 Materials and methods “Jugal, rostrally expanded to overlap most of the maxilla laterally” has been eliminated. Changes can be traced in the modified matrix PIN 2585/4a has been referred to Batrachognathus volans by available as Supplementary Information. As no lagerpetid was add- several authors (e.g., Bakhurina & Unwin 1995; Unwin & Bakhu- ed as outgroup in this analysis, the topology at the base of the strict rina 2000; Costa et al. 2013; Hone 2020; Wei et al. 2021), but the consensus tree differs from that obtained by Ezcurra et al. (2020). diagnostic features of the species, and genus by monotypy, listed by However, this is irrelevant to the purpose of this paper. Hone (2020: 1688) have not been observed in this specimen. The The Nexus data matrix used by Wei et al. (2021: https:// referral of PIN 2585/4A to Batrachognathus volans must be consid- dfzljdn9uc3pi.cloudfront.net/2021/11161/1/Wei_et_al_2021_ ered provisional and essentially based on the same stratigraphic and Sinomacrops_bondei.nex) to produce the strict consensus tree of geographical provenance as the holotype. Wei et al. (2021: fig. 7) has also been modified according to the Details of the two specimens PIN 52-2 and PIN 2585/4A results of this paper (see SI) and then used to perform a parsi- have been observed in the published figures (Wellnhofer 1991: 101; mony-based phylogenetic analysis in TNT 1.5 (Goloboff & Cata- Bakhurina & Unwin 1995: fig. 8; Unwin & Bakhurina 2000: fig. lano 2016) under the same conditions specified by Wei et al. (2021). 21.3; Hone 2020: fig. 5; and Unwin 2006: fig. 5.3) as well as some Character states regarding the jugal, maxilla, premaxilla and skull high resolution photographs showing different views and taken fenestrae in the anurognathid taxa have been coded consequently. with different illumination. As the author of this paper is not the No new characters have been added and characters and character author of those photographs and did not get a formal permission states have not been controlled for correction if wrong, ambig- to publish them, only the drawings traced based on them are re- uous, poorly-formulated and redundant. My purpose was not the ported here. Information about the other mentioned specimens improvement of the data matrix, but only checking the impact of derives only from the literature. the changes in character codings according to the results of this pa- The “rostrum” is considered here the part of the skull an- per on the relationships between the Anurognathidae and the other terior to the orbit and “rostral” refers to this part of the skull. As pterosaurs. Anyway, some problems have been encountered. For the rostral portion of the anurognathid skull is unusually short, I example, the numbers of the characters mentioned in the text of preferred to use here “anterior” and “posterior” instead of “ros- Wei et al. (2021) do not correspond with the character numeration tral” and “caudal” in the description of the polarity of the cranial in the matrix. Wei et al. (2021: 27) mention “Character 95 (1). Jugal, bones to avoid confusion. External naris and narial fenestra are lacrimal process, subvertical”, but this is character 91 in the Nexus considered as synonyms. matrix. Character 21 presents three states (0 to 2) in its definition, The first phylogenetic definition of the Anurognathidae but the same character has four character states in the matrix (0 to was published by Unwin (2003: 176): Anurognathus ammoni, Batra- 3). Six characters of the matrix are constant and 52 variable charac- chognathus volans, their most recent common ancestor, and all its de- ters are parsimony-uninformative; thus, the parsimony-informative scendants. This definition has priority and there would appear to be characters are 320 out of 378. 26 Dalla Vecchia F.M. Previous reconstructions of the and Batrachognathus volans. Andres et al. (2010: 187) anterior skull fenestrae in the were the first to opt for this coding because they ob- anurognathids served that the thin ascending process of the maxilla of Bennett’s reconstruction is doubtfully present in The first anurognathid to be described by Anurognathus ammoni and absent in specimens PIN Döderlein (1923), Anurognathus ammoni, had an in- 13 and PIN 2585/4 of Batrachognathus volans and in a complete and poorly preserved skull, which was specimen (CAGS–Z070/CAGS–IG-02-81) figured tentatively reconstructed by Wellnhofer (1975b: by Ji & Yuan (2002) and referred to Jeholopterus. The fig. 38b) as that of the non-monofenestratan pter- nasoantorbital fenestra was coded as present in the osaurs, with distinct narial, antorbital and orbital four anurognathid species included in the analysis fenestrae, the rounded antorbital fenestra being the by Vidović & Martill (2018). The nasoantorbital largest. fenestra was coded as present in Jeholopterus ningchen- The skull of Anurognathus ammoni was re- gensis, Anurognathus ammoni and Batrachognathus volans constructed in a different way by Bennett (2007), also in the phylogenetic analysis by Britt et al. (2018) based on a second, complete and articulated speci- and Dalla Vecchia (2019); although this coding was men of this pterosaur, which is represented by slab not explained by these authors, they independently and counterslab kept separately in two private col- did it for the same reason adduced by Andres et al. lections in Germany. The skull of this specimen is (2010). exposed in dorsal view flattened dorsoventrally; it Hone (2020: 1681) considered as “heterodox is relatively complete and articulated but not well scorings of character states” the identification of preserved (Bennett 2007: fig. 3). The incomplete, a “confluent nasoantorbital” fenestra by Andres et crushed and broken cheek bones of that skull are al. (2010, 2014), ignored the scorings by Vidović & difficult to interpret; Bennett (2007: 380) tentatively Martill (2018), Britt et al. (2018) and Dalla Vecchia identified a left maxilla in dorsal view overlapped (2019), and preferred the reconstruction by Bennett laterally by the jugal and preserving a row of three (2007) of narrow and distinct narial and antorbital teeth in its midline (Bennett 2007: fig. 3). A slen- fenestrae. With Hone’s (2020: 1683) words: “Nar- der and sigmoid skeletal element that “arises from is [in anurognathids] is dorsoventrally tall though the anterior end of the body of the jugal in con- anteroposteriorly short and sits at the very anterior tact with the maxilla” is reported as the “ascend- margin of the skull. Antorbital fenestra similar in ing process of the jugal” and “ascending process appearance and separated from the naris anterior- of the jugal plus the lacrimal and nasal” in the text ly and the orbit posteriorly by only thin splints of (pp. 381 and 382, respectively), while it is identi- bone. Orbit is extremely large and occupies approx- fied as the “jugal/nasal process” in figure 3. Based imately half of the lateral face of the skull”. mainly on this second specimen, Bennett (2007: fig. Wei et al. (2021: 11) also supported the pres- 4) reconstructed the skull of Anurognathus ammoni ence of a nasoantorbital fenestra in the anurog- with an anteriorly extended jugal that overlays the nathids for the same reason as Andres et al. (2010). maxilla laterally fusing with it, with the exclusion of Following Bennett (2007), they considered the the anteriormost portion of the maxilla that articu- process separating the orbit from the single rostral lates with the premaxilla. This anteriormost portion fenestra as the lacrimal process of the jugal, bone has a thin ascending process separating the narrow which is supposed to be laterally fused with not vis- antorbital fenestra from the external naris. In this ible sutures to the maxilla to form a “jugo-maxilla” reconstruction, the ascending process of the jugal making most of the jaw. separates the enormous orbit from the narrow an- torbital fenestra. Andres (2010) coded as “unknown” (?) the Systematic Palaeontology presence/absence of a nasoantorbital fenestra in Batrachognathus volans and others anurognathids in Diapsida Osborn, 1903 the matrix of his analysis, but Andres et al. (2010, Pterosauria Kaup, 1834 2014) coded “confluent external naris and antor- Anurognathidae Kuhn, 1967 (ex Anurognathinae bital fenestra” as present in Jeholopterus ningchengensis Nopcsa, 1928) sensu Hone (2020) Orbitoantorbital fenestra in the anurognathids 27 Batrachognathus Ryabinin, 1948 jacent to the maxilla is triradiate. Its anterior ramus is very short and clearly overlapped the tapering Batrachognathus volans Ryabinin, 1948 posterolateral process of the left premaxilla dor- sally. The dorsal ascending ramus is nearly vertical, Description and comparison long and narrow. Its dorsal end is covered by other The skull and mandible of the holotype of B. skeletal elements; it appears to expand slightly me- volans (PIN 52-2; Fig. 1) are dorsoventrally crushed diolaterally toward the apical part. The posterior ra- and flattened. They are exposed in ventral view mus is by far the largest of the three; it is strap-like, and slightly disarticulated, with the mandible partly slightly tapering posteriorly and slightly arched dor- overlapping the upper jaw. Despite this, the anterior sally and medially. The dorsomedial margin of the cranial elements and the mandible of PIN 52-2 are distal portion of this ramus is overlapped by the left by far the best preserved among the anurognathids. mandibular ramus; the distal end appears to be cut This skull is better preserved than that of the sec- and blunt. Ventrally, along the lateral margin, this ond specimen of Anurognathus ammoni used by Ben- longest ramus bears a row of teeth (eight teeth, ac- nett (2007) for the reconstruction of the skull of cording to Ryabinin 1948). The row extends at least that species, taken since then as reference for the along 59% of the ventral margin of the bone. Teeth anurognathid skull morphology. With respect to the appear to be similar in shape and size to the pre- second specimen of Anurognathus ammoni, the main maxillary ones. This triradiate element includes the anterior skull elements of PIN 52-2 are better pre- left maxilla because of its spatial relationships with served and exposed and do not need to be inter- the premaxilla and the fact that the only tooth bear- preted. ing upper jaw bones are the premaxilla and maxilla The description and review of the entire skull in reptiles (Romer 1956). In fact, Ryabinin (1948: of PIN 52-2 is beyond the scope of this paper; this fig. 1) had identified this element as the left max- will hopefully be done by someone else with the illa. The short anterior ramus is the premaxillary specimen at hand. I will focus here only on the an- process of the maxilla and the vertical ramus cor- terior skull fenestrae and the skeletal elements sur- responds to the ascending process of the maxilla. rounding them. When articulated, the premaxillary process of the The premaxillae of PIN 52-2 are strongly ap- maxilla and the maxillary process of the premax- pressed against each other but not fused (the suture illa form a bar that borders ventrally the external is still visible) forming a T-shaped bone as charac- naris. The posterior ramus is plausibly formed by teristic of anurognathids. Each premaxilla is com- the fusion of the jugal process of the maxilla with posed of a small main body and two processes. The the maxillary process of the jugal as occurs in other posterodorsal process is relatively broad at the base, pterosaurs and probably includes also the body of but it tapers distally; it is mostly concealed by the the jugal, as suggested also by its length with respect overlap and juxtaposition of other skull elements, to the mandibular ramus. Therefore, the element probably including the vomers. The posterolateral is a maxillojugal. In Caelestiventus hanseni (see Britt process has a trapezoidal outline, with a tapering et al. 2018: fig. 3h-i and personal observation) and posterior end where a corresponding process of the Seazzadactylus venieri (see Dalla Vecchia 2019: fig. 7A) adjacent bone overlapped dorsally. The ventral mar- the jugal overlaps the jugal process of the maxilla gin of each premaxilla bears slender and recurved laterally, forming the bar that borders the antorbital teeth, which are in number of three according to fenestra ventrally. The jugal may remain unfused to Ryabinin (1948). The bones that are posterolaterally the maxilla, as it is the case of Seazzadactylus venieri, adjacent to the joint premaxillae are slightly shifted or be fused to form a maxillojugal as in Caelestiven- posteriorly from their anatomical articulation with tus hanseni. In the latter taxon, the suture between the premaxillae; only the anterior end is visible of the jugal and the maxilla is obliterated, whereas the the right one, because the rest is covered by some premaxillae are not fused to the maxillae and the limb bones, whereas the left one is better exposed. dentaries are not fused at the symphysis, like in PIN The latter shows its medioventral aspect because of 52-2. The anterior extent of the overlap of the ju- crushing and is partially overlapped posteriorly by gal on the maxilla, if visible, may not be established the left mandibular ramus. This large element ad- in PIN 52-2, because the posterodorsal part of the 28 Dalla Vecchia F.M. Fig. 1 - Batrachognathus volans, PIN 52-2, holotype, skull and lower jaw in ventral view. Interpretative drawing. The clearly identifiable portion of the premaxilla is marked in fuchsia, the maxilloju- gals in green, the palatine/ ectopterygoids in blue, and the mandible in orange. The other bones are in gray co- lour, teeth in dark gray and patches of soft tissue are in pale yellow. The margins of the bones or portions of bone that are preserved only as an impression are repre- sented by dashed lines. Abbreviations: apm, ascend- ing process of the maxilla; en, external naris (narial fe- nestra); l, left; lbl, long bone of the limbs; mar, mandibu- lar ramus; mx, maxilla; mxj, maxillojugal; pl, palatine/ ectopterygoid; pmx, pre- maxillae; ppm, premaxillary process of the maxilla; q, quadrate; r, right; th, tooth. Scale bar = 10 mm. maxillojugal is concealed by the overlapping man- and the lacrimal, separating the orbit and the an- dibular ramus. Regardless, there is no evidence that torbital fenestra. Only a single large opening occurs the jugal extended as far anteriorly as hypothesized posterior to the ascending process of the triradiate by Bennett (2007) for Anurognathus. The jugal does bone. not totally cover the maxilla laterally, because the Unfortunately, the crushed, slightly disar- teeth are set on the lateral edge of the maxilloju- ticulated, and only partly preserved posterior part gal, which is consequently formed by the maxilla of the skull gives no information about the tem- only. Therefore, there is no reason to assume that poral fenestrae and their separation from the orbit. the single ascending process of the maxillojugal is A postorbital cannot be identified and the rod-like an anteriorly migrated lacrimal process of the ju- straight bone contacting the posterior part of the gal. In all non-monofenestratan pterosaurs the bar left mandibular ramus is probably the left quadrate. separating the external naris from the immediately Other unambiguously identifiable skull bones posterior opening (the antorbital fenestra) is the as- in PIN 52-2 are the filiform and forked elements cending process of maxilla. Instead, in PIN 52-2 characteristic of the anurognathid palate, which there is no evidence of a jugal-lacrimal bar, formed have been identified as the palatine/ectopterygoids by the anterodorsal (lacrimal) process of the jugal by Bennett (2007: fig. 5). Despite the crushing of Orbitoantorbital fenestra in the anurognathids 29 the skull, the palatines/ectopterygoids are still lo- cated in correspondence to their medial articulation with the maxilla just posterior to the ascending pro- cess. The skull of PIN 2585/4A (Fig. 2) is articu- lated, probably exposed in dorsal view and flattened dorsoventrally. The skull overlaps the mandible, but the left mandibular ramus is slightly rotated to the right with respect to the upper jaw and is thus not totally concealed by the upper jaw and is partially visible in dorsal view. Most of the posterior part of the skull probably remained on the counterslab or split away and is not preserved on this slab. Be- cause of the articulation of the cranial bones, the strong crushing, and the partial overlapping of upper and lower jaws, it is difficult to distinguish between mandible and premaxillae + maxillojugals. However, the anterior skull fenestrae are unambig- uously identifiable. It is evident that there are only Fig. 2 - Batrachognathus volans, PIN 2585/4A, skull and lower jaw in two large anterior fenestrae in this skull: a rostral dorsal view. Interpretative drawing. Bones that are probably D-shaped narial fenestra bordered posteriorly by the preserved only as an impression are represented by dashed lines. The two anterior skull openings are evidenced with ascending process of the maxilla and an enormous grey colour. posterior fenestra which is undivided and reaches Abbreviations: apmx, ascending process of the maxilla; en, external naris (narial fenestra); lbl, long bone of the limbs; very posteriorly in the skull. There is no evidence lso, lacrimal or supraorbital; ma, mandible; mxj, maxilloju- of a jugal-lacrimal bar in the lateral side of the skull. gal; oao, orbitoantorbital fenestra; pal, palatine/ectoptery- Missing the posterior part of the skull, there is no in- goid; po, postorbital; pt, pterygoid; pmx; premaxilla; q, quadrate; and th, tooth. Scale bar = 10 mm. formation on the temporal openings and the poste- rior end of the orbit. Two slender and straight bone 2 and 5A; Zhou 2014: fig. 5). Alternatively, it could impressions may represent the right postorbital and be the lacrimal, whose anatomical position is at the right quadrate given their position, but nothing can anterior margin of the orbit and was located above be said about their morphology. The ascending pro- the maxilla in basal reptiles (Romer 1956), or the su- cess of the left maxilla has a lanceolate shape and praorbital. The lacrimal or supraorbital would have appears to be slightly disarticulated from a slender migrated anteriorly, overlapping the ascending pro- element that probably overlapped it posteriorly, de- cess of the maxilla posteriorly and medially. scending from the median dorsal bar of the skull. Unlike Anurognathus (see Bennett 2007), there Because of crushing, this latter element partly over- is no evidence of the sclerotic ring elements in the laps a ramus of the forked palatine/ectopterygoid; orbits of the two Batrachognathus specimens. similarly, on the right side, a ramus of the right pal- atine/ectopterygoid is contiguous with the ascend- ing process of the maxilla. The element posterior Discussion to the ascending process is possibly preserved also in the right side of the skull, slightly disarticulated The anterior skull fenestrae in the from its anatomical position (Fig. 2). That element anurognathid skull may be the nasal as suggested for Anurognathus by In his reconstruction of the skull of Anuro- Wellnhofer (1975b: fig. 38b), but the nasal is usual- gnathus ammoni, Bennett (2007) drew a thin ascend- ly placed anterior or anterodorsal to the ascending ing process of the maxilla separating the external process of the maxilla in the reconstructions of the naris from the antorbital fenestra. As seen above, early pterosaurs (Wellnhofer 1978: fig. 2; Sangster evidence of this division of the rostral fenestra is 2003: fig. 2.2; Padian 2008a: fig.18; Padian 2008b: lacking in the holotype of Anurognathus ammoni, am- fig.10; Cheng et al. 2012: fig. 5; Bennett 2014: figs. biguous in the second specimen and absent in other 30 Dalla Vecchia F.M. anurognathid specimens; this has been remarked in Evolution of the anterolateral skull most of the recent papers dealing with the subject. fenestrae in the pterosaurs The specimens of Batrachognathus volans present a In the earliest non-monofenestratan ptero- D-shaped narial fenestra anterior to the ascending saurs (Fig. 3A), the external naris, antorbital fenestra process of maxilla and a single much, larger fenestra and orbit are distinct openings, which is the plesio- posterior to the ascending process of maxilla which morphic condition of the Archosauriformes (Nes- is formed by the confluence of orbit and antorbital bitt 2011). The external naris is bordered anterior- fenestra. ly and dorsally by the premaxilla, ventrally by the The features observed in the two Batrachogna- premaxillary process of the maxilla and in some thus specimens occur also in an articulated anurog- taxa also by the maxillary process of the premaxilla nathid skull from China figured in Ji & Yuan (2002: (which never extends beyond the anterior half of figs 1 and 2A) and reported as CAGS–Z070 by the margin, anyway), posteriorly by the ascending Yang et al. (2019: fig. 1b, SI fig. 3A) but referred process of the maxilla and posterodorsally by the to as CAGS–IG-02-81 by Hone (2020: fig. 3B) and nasal. The antorbital fenestra is bordered anteriorly Wei et al. (2021: fig. 13). This specimen is figured by the ascending process of the maxilla, ventrally by also in Unwin (2006: fig. 11.1) and Lü et al. (2006: the bar composed of the jugal process of the maxil- figs 3.5 and 3.6). It was mentioned by Andres et al. la and the maxillary process of the jugal, posteriorly (2010) regarding the absence of a vertical bar divid- by the bar formed to a variable extent by the anter- ing the rostral skull opening. It has been referred odorsal (lacrimal) process of the jugal and the lac- to Jeholopterus ningchengensis by Lü et al. (2006), Hone rimal, and dorsally by the ascending process of the (2020) and Wei et al. (2021), but is identified as an maxilla (anteriorly) and the lacrimal (posteriorly). In indeterminate anurognathid by Yang et al. (2019). It some taxa, like Dimorphodon macronyx, the nasal may comes from the Middle-Upper Jurassic of the Dao- have participated to the anterodorsal margin of the hugou locality of China (Yang et al. 2019). The skull antorbital fenestra (see reconstruction by Sangster of this specimen appears to be relatively well-pre- 2003: fig. 2.1A). The orbit is bordered anteriorly by served (see Lü et al. 2006: figs 3.6; Wei et al. 2021: the jugal-lacrimal bar separating it from the antor- fig.13), but it has not yet been described in detail. bital fenestra, ventrally by the jugal, posteriorly by The published photos and drawings by Lü et al. the postorbital and by the postorbital rami of jugal (2006: figs 3.5-6) and Yiang et al. (2019: SI fig. 1B), and frontal, dorsally by the frontal, and anterodor- and description and drawing by Wei et al. (2021: sally by the supraorbital. Fig. 13B) are in agreement with the observations In later non-monofenestratan pterosaurs, like made on Batrachognathus specimens: the ascending Angustinaripterus and Scaphognathus (Fig. 3B), the pre- process of maxilla separates the D-shaped external maxilla is excluded from the ventral and anterior naris from a single large fenestra made by the orbit margin of the external naris and the anterior end of plus the antorbital fenestra and is the only ascend- the opening is bordered by an anterodorsal process ing process identifiable laterally on the skull, and the of the premaxillary process of the maxilla. forked palatine/ectopterygoid join the maxilla me- In monofenestratan pterosaurs (Fig. 3C), the dially in correspondence of the base of the ascend- external naris and antorbital fenestra are confluent ing process. Also in this case, the posterior portion forming a nasoantorbital fenestra because of the of the skull appears to be poorly preserved as in the disappearance of the ascending process of maxil- two Batrachognathus specimens and the sclerotic ring la, apparently starting from a maxillary morphology elements may be missing. like that in Figure 3B (i.e., that bearing an anter- Given the peculiar skull shape shared by all odorsal process on the premaxillary process). The anurognathids (excluded possibly ‘Dimorphodon’ elongation of the rostrum anterior to the antor- weintraubi) and the conservatism of anurognathid bital fenestra and the enlargement of the external anatomy in general (Hone 2020), it is likely that naris produced a corresponding elongation of the the orbitoantorbital fenestra is synapomorphic of narial contribution to this composite fenestra. The all anurognathids (with the possible exception of nasal still sends a long anteroventral process into ‘Dimorphodon’ weintraubi), unless evidence appears to the nasoantorbital fenestra in basal Monofenestrata the contrary. (e.g., Darwinopterus; Lü et al. 2010a) and basal Pter- Orbitoantorbital fenestra in the anurognathids 31 Fig. 3 - Skull reconstructions and fenestration patterns of early- diverging non-monofenestratan, monofenestratan and anurognathid pterosaurs. A) The basal Triassic pterosaur Seazzadactylus venieri; B) the basal Late Jurassic pterosaur Scaphognathus crassisrostris; C) the basal monofenestratan Dar- winopterus modularis; D) the anurognathid Batrachognathus vo- lans. A is from Dalla Vecchia (2019), modified; B is redrawn from Wellnhofer (1975b) and modified according to Bennett (2007); C is redrawn and modified from Lü et al. (2010a); D is based on the reconstruction of the skull of Anurognathus by Bennett (2007) and modified according to morphology of the anterior part of the skull shown in the two specimens of Batrachognathus volans. In figure D, the shape of the frontal and the posterior part of the skull, included the temporal fenestrae, are those hypothesized for Anurognathus by Ben- nett (2007), because they cannot be reliably reconstructed in Batrachognathus. As all anurognathid skulls are flattened and quadrates and postorbitals are poorly or not preserved, the depth of the skull is hypothetical in figure D. Also mandibu- lar teeth are hypothetical in figure D. Abbreviations: eaof, antorbital fenestra; en, external naris (narial fenestra); fr, frontal, j, jugal; lac, lacrimal; ltf, lower temporal fenestra; mx, maxilla; na, nasal; nao, nasoantorbital fenestra; or, orbit (orbitoantorbital fenestra in Batrachogna- thus volans); pmx, premaxilla; so, supraorbital, utf, upper tem- poral fenestra. Not drawn to scale. odactyloidea (e.g., Germanodactylus and Pterodactylus; Wellnhofer 1970), which nearly disappeared in later pterodactyloids (e.g. Pteranodon; Bennett 2001). In the anurognathids, the modification of the snout is the opposite to that of the Monofen- estrata. The shortening of the rostrum is obtained by the extreme anteroposterior curtailment of the premaxillary body and of the premaxillary process of maxilla. The jaws arched laterally, forming the unique, short and broad anurognathid skull mor- phology. The external naris is resultedly deeper than long and bordered ventrally mainly by the premax- illa, unlike all other pterosaurs. The bar separating the orbit and the antorbital fenestra disappeared, forming a large orbitoantorbital fenestra. This was attained by the loss of the anterodorsal (lacrimal) postorbital articulates directly with the body of the process of the jugal; the lacrimal was also lost or jugal. Therefore, the jugal may be simply a horizon- migrated anteriorly following the verticalization tal anteroposterior bar overlapping the jugal process of the ascending process of maxilla. Alternatively, of the maxilla to an unknown anterior extent and the lacrimal was lost with the anterodorsal (lacri- bordering the orbitoantorbital fenestra. The loss of mal) process of the jugal and the bone posterior to the jugal-lacrimal bar would have allowed for the the ascending process of maxilla is the anteriorly expansion of the orbital space leading to the devel- shifted supraorbital. It is worth note that, according opment of an enormous eye. to the reconstruction of the skull of Anurognathus ammoni by Bennett (2007) used as the base for the Phylogenetic analysis tentative reconstruction of the Batrachognathus volans The data matrix of Dalla Vecchia (2019) (see skull of Figure 3D, the postorbital process of the Materials and Methods section) has been modified jugal is also lost in the anurognathid skull and the for the characters regarding the external naris, an- 32 Dalla Vecchia F.M. Fig. 4 - Phylogenetic position of the Anurognathidae based on the modified matrix of Dalla Vecchia (2019). Strict consensus tree of nine most parsimonious trees as computed in PAUP 4.0b10 (length = 281 steps, CI = 0.5587, RI = 0.6923). Legend: 1, Pterosauria; 2, Macronychoptera; 3, Lon- chognatha; 4, Novialoidea; 5, Caelidracones; 6, Breviquar- tossa; 7, Monofenestrata; 8, Pterodactyloidea; 9, Anu- rognathidae (sensu Hone 2020); 10, Anurognathidae (sensu Unwin 2003). torbital fenestra, orbit, premaxilla, maxilla and jugal 378 (10%) have been revised based on the results of in the anurognathids, according with the results of this paper in the matrix by Wei et al. (2021), which is this paper. In the resulting strict consensus tree of the most recently published matrix, includes seven nine most parsimonious trees, the Anurognathidae anurognathid taxa and was specifically constructed are the sister group of Sordes pilosus + Breviquartos- to investigate the anurognathid relationships (both sa (Fig. 4). With respect to the strict consensus tree internal and with other pterosaur clades). Informa- of Dalla Vecchia (2019), the trichotomy between tion about the modified character states is includ- Sordes pilosus, Scaphognathus crassirostris + Rhampho- ed in the Supplementary Information. The goal rhynchinae and Monofenestrata is resolved, while of this revision is to investigate the impact that the relationships within the Anurognathidae are less the presence of an orbitoantorbital fenestra in the resolved and Dendrorhynchoides curvidentatus results anurognathid skull has in the strict consensus tree more basal than Anurognathus ammoni and Batracho- produced running the modified matrix. The goal gnathus volans. The clade Dendrorhynchoides curvidenta- is not to produce a phylogenetic hypothesis alter- tus + Anurognathidae sensu Unwin (2003) corre- native to that of Wei et al. (2021). Therefore, the sponds to the Asiaticognathidae of Kellner (2003). presence of an orbitoantorbital fenestra, a distinct However, the matrix by Dalla Vecchia (2019) narial fenestra, an ascending process of maxilla, and was made to investigate the relationships of the ear- the absence of a jugal-lacrimal bar have been coded liest (Triassic) pterosaurs, not those of the anuro- for all anurognathid taxa under the assumption of gnathids (it includes only Anurognathus ammoni, Ba- the conservatism of anurognathid cranial anatomy trachognathus volans, Dendrorhynchoides curvidentatus and postulated by Hone (2020), even when the character Jeholopterus ningchengensis). Thus, 38 characters out of states are not actually observed in some taxa. Orbitoantorbital fenestra in the anurognathids 33 Fig. 5 - Phylogenetic position of the Anurognathidae based on the modified matrix of Wei et al. (2021). Strict consensus tree of two most parsimonious trees as computed in TNT 1.5 (length = 1110 steps, CI = 0.4541 [CI excluding uninformative characters = 0.4250], RI = 0.6735. Legend: 1, Monofenestrata; 2, Darwinoptera; 3, Pterodactyloidea; 4, Anurognathidae sensu Hone (2020); 5, Anurognathidae sensu Unwin (2003). The formulation of most of the characters pilosus + Fenghuangopterus lii + (Jianchangnathus robustus regarding external naris, antorbital fenestra, orbit, + Scaphognathus crassirostris). The clade Anurognathi- nasoantorbital fenestra, maxilla and jugal done by dae sensu Hone 2020 + (Sordes pilosus + Fenghuan- Wei et al. (2021) makes them “inapplicable” for gopterus lii + (Jianchangnathus robustus + Scaphognathus anurognathids. This is somewhat problematic, be- crassirostris)) is the sister group of the Monofene- cause “inapplicable” characters are not treated strata. as such by TNT, but are practically considered as Thus, the presence of an orbitoantorbital codifiable but “unknown”. Furthermore, the states fenestra in the anurognathids (excluding ‘Dimorpho- of characters regarding the nasal and the lacrimal don’ weintraubi), based on the identification of the have been considered as unknown because the maxilla instead of the jugal as the bone bearing presence and position of these elements is still to the ascending process separating the two anterior be confirmed in the anurognathids. This decreased skull fenestrae, has an impact in the position of the the number of codified characters for the anurog- Anurognathidae in the cladogram. nathids. Because the Monofenestrata are defined as a The analysis found two most parsimonious synapomorphy-based clade by Andres et al. (2014: trees like that by Wei et al. (2021); in the resulting SI), i.e. “Monofenestrata after Lü et al. (2010)..., the strict consensus tree (Fig. 5; five steps shorter than most inclusive clade exhibiting confluent external that by Wei et al. 2021), the Anurognathidae sensu naris and antorbital fenestra synapomorphic with Hone 2020 are the sister group of the clade Sordes that in Pterodactylus antiquus Soemmerring 1812”, 34 Dalla Vecchia F.M. the presence of an orbitoantorbital fenestra would Anurognathus by Bennett 2007: fig. 4A) or about as place Anurognathidae outside Monofenestrata by long as broad (as in Batrachognathus PIN 2585/4A) definition. and the jaws are broadly arching laterally (Wei et al. The single specimen of ‘Dimorphodon’ wein- 2021: figs. 8-9). Height is the shortest skull dimen- traubi preserves only the portion of the skull pos- sion, thus the skulls of these pterosaurs are usually terior to the orbits, which has not been described preserved in dorsoventral view and are reminiscent yet (Clark et al. 1994, 1998). Thus, we do not know of those of frogs (the name Batrachognathus means whether this taxon had an ‘anurognathid’ skull or “frog jaw”). Unlike other pterosaurs, the skull is not and whether the ‘anurognathid’ skull is a feature comparatively small with respect to body size, espe- of the Anurognathidae sensu Hone 2020 or of a cially in largest and probably more mature individ- less inclusive clade. uals. This is evident in the holotypes of Jeholopterus The matrix presents many unknown charac- ningchengensis (see Wang et al. 2002) and Vesperoptery- ter state codings that could be codified after an ad- lus lamadongensis (see Lü et al. 2018); the small size of equate study of the specimens of Batrachognathus vo- the skull is independent from the extreme reduction lans, Jeholopterus ningchengensis, other still unpublished of the rostrum. Chinese anurognathid specimens, and ‘Dimorphodon’ The unique fenestration pattern, with an weintraubi. enormous orbitoantorbital and a much smaller and D-like external naris, is further evidence of the dis- The unique anurognathid bauplan tinction of the skull bauplan of this clade from that There is a general agreement that the Anuro- of all other pterosaurs. gnathidae are a distinct and peculiar group among The structure of the palate is also distinc- the pterosaurs (Bennett 2007; Hone 2020; Wei et al. tive of the anurognathids (Bakhurina 1988; Ben- 2021). Anurognathids share with other non-ptero- nett 2007). The palatines/ectopterygoids, branched dactyloid pterosaurs a digit V of the pes composed vomers and quadrate ramus of the pterygoid and/ of two elongated phalanges (with the exclusion of or pterygoid ramus of the quadrate, according to Campylognathoides, which has a somewhat reduced the reconstruction by Bennett (2007), are slender, pedal digit V; Padian 2008b) and a short wing met- splint- or rod-shaped bones with a peculiar mor- acarpal. Digit V of the pes is reduced in all ptero- phology that is not observed in any other pterosaur dactyloids and the wing metacarpal (metacarpal IV) (Wellnhofer 1978; Ősi et al. 2010). In particular, the is proportionally more elongated than in non-pter- palatine/ectopterygoid is characteristically forked. odactyloid pterosaurs (“pes digit V with a single These palatal elements are more or less preserved phalanx or entirely absent” and “wing metacarpal in Anurognathus ammoni, Batrachognathus volans, Den- at least 80% the length of the humerus” are syn- drorhynchoides curvidentatus, Jeholopterus ningchengensis, apomorphies of the Pterodactyloidea according to and Luopterus mutoudengensis (fide Lü & Hone 2012). Unwin 2003). The “elongation of the metacarpus is Other features that have been considered di- the diagnostic apomorphy of the Pterodactyloidea” agnostic of the anurognathids are somewhat ambig- according to Andres et al. (2014: 1015) and “referral uous, not universally accepted or occur also in other to the Pterodactyloidea requires only that the wing pterosaur taxa. metacarpal be longer than 80% of the humeral Cervical ribs and vertebrae. Mid-cervical ribs (i.e., length” (Andres et al. 2014: Supplemental Experi- ribs of cervical vertebrae 3-7) are absent in juve- mental Procedures). nile Anurognathus ammoni (the first seven cervicals Anurognathid diagnostic features have been lack ribs; Bennett 2007) and in the only specimen reviewed and discussed by Bennett (2007), Andres of Vesperopterylus lamadongensis (no cervical ribs are et al. (2010), and Hone (2020). associated with the preserved cervical vertebrae; Lü Skull and mandible. The most apparent fea- et al. 2018); the complete loss of mid-cervical ribs is a ture of the anurognathid bauplan is the short and synapomorphy of Anurognathus ammoni and Vesperop- broad skull, which has a hemicircular to nearly cir- terylus lamadongensis according to Wei et al. (2021). cular outline in dorsal and ventral views and a very Dendrorhynchoides curvidentatus may lack cervical ribs short rostrum with T-shaped premaxillae. The skull (Unwin et al. 2000: fig. 2); Ji & Ji (1998) and Unwin is broader than long (as in the reconstruction of et al. (2000) do not mention them in their descrip- Orbitoantorbital fenestra in the anurognathids 35 tion of the specimen. “Short and slender cervical is reminiscent of the holotype synsacrum of Mesa- ribs” are reported in Jeholopterus ningchengensis (see dactylus ornithosphyos from the Upper Jurassic of the Wang et al. 2002: 227). Cervical ribs are not men- USA, suggesting that the latter is an anurognathid. tioned in the description of Sinomacrops bondei by The synsacrum of Mesadactylus ornithosphyos is com- Wei et al. (2021). No information is available about posed of seven vertebrae. According to Wei et al. the cervical ribs of Batrachognathus volans and Luopte- (2021: 12), the sacrum of Sinomacrops bondei also pos- rus mutoudengensis. In all these taxa, the cervical ver- sesses ribs/transverse processes of the first sacral tebrae are poorly preserved, but they appear to be vertebra that are strongly inclined posteriorly, like short and stocky. Also the preserved cervicals of those of Mesadactylus ornithosphyos; the long sacrum ‘Dimorphodon’ weintraubi are short and stocky (Clark suggests the presence of many sacral vertebrae in et al. 1998). “Cervical vertebrae short and robust” the Chinese species. According to Hone (2020), the is a diagnostic feature of the Family Anurognathi- presence of a supraneural plate on the sacrum of dae according to Bennett (2007: 395). No vertebral Mesadactylus ornithosphyos suggests it may belong to feature is considered characteristic of the Anuro- a pterodactyloid (the original referral of Mesadacty- gnathidae by Hone (2020). lus ornithosphyos by Jensen & Padian 1989). However, Earliest non-monofenestratan pterosaurs the sacrum is poorly preserved in both specimens have long and filiform cervical ribs at least up to of Anurognathus ammoni (see Bennett 2007: figs. 2 two vertebral centra in length (Padian 2008a, b; Dal- and 9), Sinomacrops bondei (see Wei et al. (2021: fig. la Vecchia 2014). The first eight cervicals of the ju- 5), and all other anurognathids. The sacrum of the venile Scaphognathus crassirostris specimens lack ribs, rhamphorhynchid Dorygnathus banthensis (see Padian whereas cervicals 4-8 bear small cervical ribs in the 2008a: fig. 19B) appears to be similar to that of Me- adult (Bennett 2014); the cervical ribs are short also sadactylus ornithosphyos (the ribs/transverse processes in Rhamphorhynchus muensteri (see Wellnhofer 1975a). of the first sacral are inclined backward), although it All these non-monofenestratan pterosaurs have is composed of only four vertebrae. The same shape mid-cervical vertebrae that are stockier than those and orientation of the ribs/transverse processes of of monofenestratan pterosaurs. the first sacral vertebra and the similar number of The basal monofenestratan Darwinopterus sacral vertebrae might imply a close relationship of modularis seems to lack cervical ribs and has cervical Anurognathus ammoni and Dorygnathus banthensis; the vertebrae that are much more elongated than those shape and orientation of the ribs/transverse pro- of the anurognathids (Lü et al. 2010a). “Cervical cesses of the first sacral vertebra might support a vertebrae more elongated than in other non-pter- relationship of Anurognathus ammoni and Dorygnathus odactyloids” and “cervical ribs reduced” are syn- banthensis with Mesadactylus ornithosphyos, or it may apomorphies of the Wukongopteridae according just be an homoplastic feature. This aspect needs to Wang et al. (2010). Mid-cervical ribs are absent further investigation. or reduced in the Pterodactyloidea and mid-cervi- Short tail - “Short tail” is a diagnostic feature cal vertebrae are in general more elongated than of the Family Anurognathidae according to Ben- in non-monofenestratan pterosaurs (Kellner 2003; nett (2007: 395), while it is not considered as such Unwin 2003). by Hone (2020). The tails of the anurognathids are The morphology of the cervical segment of comparatively shorter than those of other non-pter- the vertebral column of the anurognathids, with odactyloid pterosaurs (Dalla Vecchia 2002; Lü et al. reduced or absent cervical ribs and comparatively 2010a; Lü & Hone 2012), but the relative tail size short and stocky cervical vertebrae, is more remi- is somewhat variable within the clade. The longest niscent of the condition in Scaphognathus crassirostris anurognathid tail gets close in relative size to that and Rhamphorhynchus muensteri than to that of other of the basal monofenestratan Douzhanopterus zhengi pterosaurs. (see Wang et al. 2017) and the shortest overlaps the Sacrum. Bennett (2007) identified a compara- pterodactyloid tail length range (Lü & Hone 2012). tively high number (four or five) of sacral vertebrae Hone (2020) reports from eight to ?20 caudals in in Anurognathus ammoni. He also noticed a configu- the anurognathids. The tail is short in Anurognathus ration of the the ribs/transverse processes of the ammoni (possibly 10 caudal vertebrae, Bennett 2007; first and second sacral vertebrae in this taxon that tail/humerus length ratio = 0.40, tail/femur length 36 Dalla Vecchia F.M. ratio = 0.50; Jiang et al. 2015) and Vesperopterylus called “Painten pro-pterodactyloid” (Tischlinger & lamadongensis (no more than 15 caudal vertebrae; Frey 2013), recovered as a basal pterodactyloid in tail/humerus length ratio = 0.44, tail/femur length the phylogenetic analysis by Wu et al. (2017), has ratio = 0.59; Lü et al. 2018). Caudal vertebrae of perhaps 17 vertebrae and bears elongated zygapo- these short-tailed taxa do not have elongated zyga- physes and haemapophyseal processes in its middle pophyses and filiform haemapophyseal processes. section; the tail/humerus and tail/femur length ra- The tail of Luopterus mutoudengensis may be propor- tios are 0.71 and 0.75, respectively. The tail/femur tionally slightly longer (possibly 15 caudal vertebrae; length ratio ranges 0.14-1.88 in pterodactyloids ac- tail/humerus length ratio = 0.67; tail/femur length cording to Lü & Hone (2012: tab. 2) with mean 0.68 ratio = 0.86; Lü & Hone 2012: fig. 3) and has fili- (N = 18), but only one (the relatively long-tailed Pte- form processes that Lü & Hone (2012) interpret- rodaustro guinazui, with at least 22 caudal vertebrae) ed as haemapophyseal processes only, while Wei et has a ratio >1 and the rest range 0.14-0.92 (mean of al. (2021) assume as probable that they represent 0.61). Excluding the basal “Painten pterodactyloid”, both elongated zygapophyses and filiform haemap- pterodactyloid caudals lack very elongated zygap- ophyseal processes, but coded their presence as ophyses and filiform haemapophyseal processes, ‘unknown’ in the matrix of the phylogenetic anal- although the zygapophyses of the mid-caudals of ysis. The tail of Sinomacrops bondei is also relatively Pterodaustro are comparatively more elongated than long (longer than the entire hind limb; Wei et al. in other pterodactyloids (Codorniú 2005). 2021). The tail is partly preserved in specimen PIN According to Bennett (2007), Anurognathus 2585/4a of Batrachognathus volans and is compara- ammoni and the pterodactyloids evolved short tails tively long (the ratio of the preserved portion of convergently because the caudal vertebrae are quite the tail/humerus length = 1.47; Costa et al. 2013) different in the two taxa, with those of Anurognathus and has elongated zygapophyses and filiform hae- ammoni being much wider transversely than anter- mapophyseal processes (Costa et al. 2013). The tail oposteriorly long. is comparatively long also in the unnamed specimen The results of the phylogenetic analyses here IVPP V16728 from the Middle-Upper Jurassic of reported support the view that the Anurognathi- China described by Jiang et al. (2015), which has dae and the Pterodactyloidea convergently acquired at least 20 vertebrae, and its tail/humerus and tail/ comparatively short tails. femur length ratios equal 1.04 and 1.50, respectively. Narrow sternum. According to Hone The caudal vertebrae of IVPP V16728 also show (2020:1684), anurognathids have a narrow sternum elongated zygapophyses and haemapophyseal pro- “approximately the length of four dorsal vertebrae, cesses. The unnamed specimen NJU–57003 from and approximately the same width as the dorsal ver- the Middle-Upper Jurassic of China reported by tebral column”, but this is not listed as a diagnos- Yang et al. (2019: fig. 2a and f, SI fig. 1A) has a long tic feature of the clade in that paper. The sternum tail, which is the proportionally longest anurog- of other pterosaurs (both non-pterodactyloid and nathid tail according to Hone (2020). The caudal pterodactyloid) is a broad plate (Wellnhofer 1978: segment of the vertebral column of NJU–57003 fig. 8; Lü et al. 2011: fig. 4; Dalla Vecchia 2014: fig. comprises 18 or19 vertebrae and is 41.8 mm long 4.2.2). Bennett (2007) did not consider the shape (the “total 418 mm” reported by Yang et al. 2019: of the sternum a diagnostic feature of the Fami- SI 2, is evidently a mistake); the tail/humerus length ly Anurognathidae, although he retained “sternum ratio is 1.45-1.52 (as humeri have different lengths) broad” as a character shared by all pterosaurs except and the tail/femur ratio is 1.96. the anurognathids (p. 395). Andres et al. (2010) con- In the basal monofenestratan Douzhanopterus sidered a narrow sternum an apomorphy of Anuro- zhengi, tail/humerus length ratio is 1.73, the tail is gnathus ammoni, because they accepted the identifi- twice the length of femur, and there are 22 caudal cation by Bakhurina & Unwin (1995) of a broad vertebrae with elongated zygapophyses and hae- sternum in Batrachognathus volans (contra Bennett mapophyseal processes (Wang et al. 2017). 2007, who apparently accepted Ryabinin’s identifi- The structure of the tail is quite variable cation). Confusingly, the sternum of Batrachognathus within the pterodactyloids (Bennett 2001; Codorniú volans has never been figured in any form, making 2005). The vertebral column of the tail of the so- verification of either shape difficult. The sternum is Orbitoantorbital fenestra in the anurognathids 37 apparently absent (not preserved or missing) in the rognathidae by Bennett (2007: 395). “Pteroid less articulated skeletons of Dendrorhynchoides curvidenta- than 1/5th of the length of the ulna” is diagnostic tus (see Unwin et al. 2000), Jeholopterus ningchengensis of the Anurognathidae according to (Hone 2020: (see Wang et al. 2002; Lü et al. 2006; Kellner et al. 1684). The pteroids appear to be comparatively 2009; Yang et al. 2019), Luopterus mutoudengensis (see small and simple rods in both specimens of Anurog- Lü & Hone 2012), Vesperopterylus lamadongensis (see nathus ammoni (see Bennett 2007), and in Dendrorhyn- Lü et al. 2018), and Sinomacrops bondei (see Wei et al. choides curvidentatus (see Unwin et al. 2000), Jeholopter- 2021). Thus, this feature needs further investigation. us ningchengensis (see Wang et al. 2002; Lü et al. 2006), Proximal end of the humeral head in dorsal view and Vesperopterylus lamadongensis (see Lü et al. 2018). nearly symmetrical. “Humerus with small, extremely However, the shape and size of the pteroid may be proximally placed deltopectoral crest” is a diagnos- ontogenetically variable. The immature specimen tic feature of the Family Anurognathidae according of Carniadactylus rosenfeldi MPUM 6009 (see Dalla to Bennett (2007: 395). According to Hone (2020: Vecchia 2018) has a comparatively small pteroid 1684), “proximal end of the humeral head in dorsal (Wild 1979: pl. 5, fig. 18); although both pteroids of view nearly symmetrical i.e. the deltopectoral crest MPUM 6009 are distally incomplete, the preserved and medial crest are similar in size” is a diagnostic part of the left one is only 16% the length of ulna. feature of the Anurognathidae. “Extremely prox- The pteroid of specimen MCSNB 3359 of Peteino- imally placed deltopectoral crest” is a somewhat saurus zambellii (which shows many features of os- ambiguous definition, as the deltopectoral crest is teological immaturity) is less than 1/5 the length located at the proximal extremity of the humerus of ulna (Dalla Vecchia 2014). The relative size of in all earliest pterosaurs (e.g., Dalla Vecchia 2009a: the pteroid seems to increase with absolute size fig. 4). Furthermore, the shape of the deltopectoral of the individuals in Rhamphorhynchus muensteri, i.e., crest appears to be somewhat variable within the small individuals have comparatively shorter pter- anurognathid species (see Wei et al. 2021: fig. 10). oids (see Wellnhofer 1975a: fig. 12a-b). The small Size and shape of the deltopectoral and “medial” size and simple shape of the anurognathid pteroids crests appear relatively similar only in Dendrorhyn- may be related to the early ontogenetic stage of the choides curvidentatus, while they are different in the specimens, or it may represent a paedomorphic fea- other taxa. There are no remarkable differences be- ture in mature individuals (Vidović & Martill 2018). tween the humeri of the anurognathids and those Notice that the pteroid/ulna length ratio may be of other early pterosaurs. somewhat misleading in representing the relative Long antebrachium. This has been consid- size of the pteroid in the different taxa, because the ered a diagnostic feature of the Anurognathidae ulna is the second longest forelimb element in the by Bennett (2007: 395), but not by Hone (2020). anurognathids (a long antebrachium was considered This elongation has not been quantified by Bennett a diagnostic feature of the anurognathids by Ben- (2007), thus the feature remains somewhat vague. nett 2007, as seen above), while it is comparatively The ulna is shorter than wing phalanx 1 in Anuro- shorter in other taxa (e.g. in Carniadactylus rosenfeldi). gnathus ammoni, Jeholopterus ningchengensis, Dendrorhyn- Very short wing metacarpal. A very short and ro- choides curvidentatus, Luopterus mutoudengensis, Vesperop- bust wing metacarpal is reported by Hone (2020) terylus lamadongensis and Sinomacrops bondei. Among as a diagnostic feature of the anurognathids based non-pterodactyloid pterosaurs, the ulna is the long- on Witton (2013) and Vidović & Martill (2018). est element of the forelimb in Preondactylus buffarinii, However, this ‘shortness’ and ‘robustness’ should Peteinosaurus zambellii, Dorygnathus banthensis, Sordes be quantified, as the wing metacarpals of non-pter- pilosus, Scaphognathus crassirostris, Jinchangopterus zhao- odactyloid pterosaurs are considered synapo- ianus, Changchengopterus pani, Pterorhynchus wellnhoferi, morpically ‘short’ compared with those of ptero- Wukongopterus lii, Darwinopterus modularis, D. robusto- dactyloids and ‘robust’ compared to metacarpals dens and D. linglongtaensis, and Kupengopterus sinensis I-III. Hone (2021: 1684) supported the shortness, (see Dalla Vecchia 2014). affirming that the wing metacarpal is “less than Short and simple pteroid. “Pteroid simple rod 1/3rd humeral length” (i.e. humerus/wing meta- no more than one-sixth of ulna length” has been carpal length ratio >3.33) in the anurognathids. In considered a diagnostic feature of the Family Anu- order of decreasing humeral length, the humerus/ 38 Dalla Vecchia F.M. wing metacarpal length ratio is actually ca. 2.75 (hu- see also Dalla Vecchia 2018: SI) and the mean is merus = 71 mm) in Vesperopterylus lamadongensis (see 1.69, but only one individual has a ratio >2; elim- Lü et al. 2018), ca. 3.26 (humerus = 62 mm) in the inating this deviant ratio, the mean becomes 1.64. holotype of Jeholopterus ningchengensis (see Wang et al. The same ratio ranges 1.55-2.15 in a sample of 40 2002), 2.91 (humerus = 32 mm) in Anurognathus am- individuals of the medium size-class (Wellnhofer moni (see Wellnhofer 1978), 2.99 (humerus = 27.8 1975b) and the mean is 1.84, with 22.5% of val- mm) in Dendrorhynchoides curvidentatus (see Unwin ues ≥2. The large size class is represented by only et al. 2000), 2.45 (humerus = 27 mm) in Luopterus two individuals and shows no difference with the mutoudengensis (see Lü & Hone 2012), and ca. 3.52 medium size-class. Therefore, there is an evident in- (humerus = 23.6 mm) in Sinomacrops bondei (see Wei crease of the ratio from the small size-class to the et al. 2021). Therefore, the ratio is quite variable, medium size-class, despite the negative allometric ranging 2.45-3.52. Apparently, this variability is not growth of the humerus with respect to body length related to body size, as the second largest and the and wing finger length (Bennett 1995: tab. 2). smallest individuals have the comparatively shortest Thus, some anurognathids have the shortest metacarpals IV (ratio >3) in the sample. The same wing metacarpals with respect to the humerus (ra- ratio ranges 1.62-2.89 in Triassic pterosaurs and is tio>3) among pterosaurs, but not all of them have highest in juvenile individuals MCSNB 8950 (2.89) such extreme proportions, overlapping the ratio and MPUM 6009 (2.50) (Dalla Vecchia 2014). The ranges of other pterosaurs. This feature is worth humerus/wing metacarpal length ratio is 2.32-2.17 further investigation; it is evident that a much larger in Dimorphodon macronyx (see Sangster 2003), 2.39 in anurognathid sample is necessary to evaluate its ac- Campylognathoides zitteli (see Padian 2008b), 1.80 in tual importance. Fenghuangopterus lii (see Lü et al. 2010b), 1.98-2.16 The wing metacarpal is shorter than the other in Scaphognathus crassirostris (see Bennett 2014), 1.74 metacarpals in the two Anurognathus ammoni speci- in Nesodactylus hesperius (see Colbert 1969), 1.96 in mens (Wellnhofer 1975b; Bennett 2007), which is Qinglongopterus guoi (see Lü et al. 2012), 2.61 in Sor- unlike the condition in all other pterosaurs, includ- des pilosus (Unwin et al. 2000), 2.63 in Orientogna- ing the other anurognathids; thus it appears to be an thus chaoyangensis (see Lü et al. 2015), 1.66-1.86 in apomorphy of Anurognathus ammoni. Changchengopterus pani (see Lü 2009; Zhou & Schoch Flexibility of joints in wing finger. Bennett (2007) 2011), 1.82 in Pterorhynchus wellnhoferi (see Czerkas & noticed that the specimens of Anurognathus ammoni, Ji 2002), 1.69 in Wukongopterus lii (see Wang et al. Dendrorhynchoides curvidentatus, and Jeholopterus ningchen- 2009), ca. 1.57 in Kunpengopterus sinensis (see Wang gensis (a sample of only four specimens when the et al. 2010), 1.64-169 in Darwinopterus modularis (see author wrote the paper) have markedly flexed inter- Lü et al. 2010a), 1.67 in D. robustodens (see Lü et al. phalangeal joints. He considered this a distinction 2011), and 1.69-1.79 in D. linglongtaensis (in a single from other pterosaurs, although he did not list it specimen; Wang et al. 2010). In a sample of seven as a diagnostic feature of the Family Anurognathi- individuals of Campylognathoides liasicus, the humer- dae. Andres et al. (2010) considered this flexion to us/wing metacarpal length ratio ranges 1.96-2.41, be a result of the disarticulation of the skeletons. with a mean of 2.21 (Padian 2008b). In a sample of Hone (2020: 1701) argued that this flexion “occurs 20 individuals of Dorygnathus banthensis, the humer- in all specimens, even when the rest of the material us/wing metacarpal length ratio ranges 1.76-2.27 is well articulated, and disarticulation of the wing with a mean of 1.99 (Padian 2008a). phalanges is rare in all other pterosaurs”. I am skep- The largest available pterosaur sample to the tical about the validity of this feature as an anuro- species level is that of Rhamphorhynchus muensteri gnathid synapomorphy until wing phalanges joints (see Wellnhofer 1975b). Rhamphorhynchus specimens of anurognathids and other non-monofenestratan have been divided into small, medium and large pterosaurs are studied as tridimensional structures size-classes, corresponding to juvenile, subadult/ and compared to look for differences. As argued by young adult and adult individuals of this species Andres et al. (2010), I also suspect that flexion may by Bennett (1995). The humerus/wing metacarpal be taphonomically biased, but due mainly to the length ratio ranges 1.27-2.92 in a sample of 26 in- dorsoventral flattening of the anurognathid body dividuals of the small size-class (Wellnhofer 1975b; rather than its disarticulation. Orbitoantorbital fenestra in the anurognathids 39 Wing phalanx 4 highly reduced or lost. The ter- sensu Hone (2020) include the Early Jurassic ‘Di- minal wing phalanx is the shortest in all non-pter- morphodon’ weintraubi which, unfortunately, does not odactyloid pterosaurs (Dalla Vecchia 2014), but it preserve much of the skull. is never reduced as in Jeholopterus ningchengensis (see With only 15 specimens reported in the lit- Wang et al. 2002: fig. 1; Kellner et al. 2009: fig. 1), erature, most of which are poorly preserved or in- Luopterus mutoudengensis (see Lü & Hone 2012: fig. 3), complete and often just preliminarily described, we and Vesperopterylus lamadongensis (see Lü et al. 2018: are far from satisfactorily knowing anurognathid fig. 2). In Anurognathus, there are no wing phalanges anatomy and its variability. Knowledge on these pe- 4 and wing phalanx 3 is thin and tapering to a blunt culiar pterosaurs will benefit from the publication point (Bennett 2007: fig. 2). Hone (2020) argued of detailed descriptions of the specimens belong- that Anurognathus ammoni had four wing phalanges ing to Batrachognathus volans, Jeholopterus ningchengen- per wing digit and that wing phalanx 4 is just un- sis, Dendrorhynchoides curvidentatus, and ‘Dimorphodon’ preserved or covered by rock in Anurognathus spec- weintraubi. imens, but he did not give any evidence to support this view. Also the only figured specimen of Dendro- Acknowledgements: I thank Fabiana Costa Rodrigues Nunes for sharing photographs of the Batrachognathus volans specimens in 2012 rhynchoides curvidentatus seems to lack wing phalan- and Andrea Cau for running the phylogenetic analysis of the matrix ges 4 (Unwin et al. 2000: fig. 2). Wing phalanx 4 within TNT 1.5. I am grateful to the reviewers Adam Fitch and Borja is present in most pterodactyloids, but is generally Holgado Palacios for their valuable comments and suggestions. Once very reduced in the azhdarchoids (Witton 2013) and again, I am indebted with Cristina Lombardo for her editorial work. was lost in Nyctosaurus and Beipiaopterus (see Bennett References 2007). The loss of the wing phalanx 4 or its ex- treme reduction is a feature convergently acquired in anurognathids and pterodactyloids. Andres B. (2010) - Systematics of the Pterosauria. PhD thesis, Yale University, New Haven, U.S.A. Andres B., Clark J.M. & Xu X. (2010) - A new rhamphorhy- Conclusions nchid pterosaur from the Upper Jurassic of Xinjiang, China, and the phylogenetic relationships of basal pter- Anurognathids are a clade of pterosaurs char- osaurs. Journal of Vertebrate Paleontology, 30(1): 163-187. Andres B. (2012) - The early evolutionary history and adaptive acterized mainly by a peculiar skull and lower jaw. radiation of the Pterosauria. Acta Geologica Sinica, 86(6): The morphology of the skull bones in relatively 1356-1365. well-preserved specimens of the anurognathid Ba- Andres B. & Myers T.S. (2013) - Lone star pterosaurs. Earth trachognathus volans shows that this pterosaur had an and Environmental Sciences Transactions of the Royal Society of enormous skull fenestra originating from the con- Edimburgh, 103(3-4): 383-398. fluence of the orbit and the antorbital fenestra and Andres B., Clark J. & Xu X. (2014) - The Earliest Pterodacty- loid and the Origin of the Group. Current Biology, 24(9): separated from the external naris by the ascending 1011-1016. process of the maxilla. Posterior to this ascending Bakhurina N.N. (1988) - On the first rhamphorhynchoid from process, the maxilla and jugal form a low bar bor- Asia: Batrachognathus volans Ryabinin 1948, from Upper dering ventrally the orbitoantorbital fenestra. The Jurassic beds of Karatau. Byuleten’ Moscowskogo Obschchest- broad fenestra hosted a very large eye. This large va Ispytatelei Prirody, Otdel Geologischeskii, 59(3): 130. [in skull opening occurs also in Jeholopterus ningchengen- Russian] Bakhurina N.N. & Unwin D.M. (1995) - A survey of ptero- sis and is plausibly a synapomorphy of the anuro- saurs from the Jurassic and Cretaceous of the former gnathids, with the possible exclusion of ‘Dimorpho- Soviet Union and Mongolia. Historical Biology, 10: 197- don’ weintraubi. The revision of the data matrices of 245. the phylogenetic analyses by Dalla Vecchia (2019) Bennett S.C. (1995) - A statistical study of Rhamphorhynchus and Wei et al. (2021) according to this new recon- from the Solnhofen Limestone of Germany: year-class- struction of the anurognathid skull shows that the es of a single large species. Journal of Paleontology, 69: 569- 580. anurognathids fall outside the Monofenestrata and Bennett S.C. (2001) - The osteology and functional morphol- are an advanced clade within the non-monofenes- ogy of the Late Cretaceous pterosaur Pteranodon. Part tratan pterosaurs that appeared in the Early Jurassic. I. General description and osteology. Palaeontographica, A In both phylogenetic analyses, the Anurognathidae 260: 1-112. 40 Dalla Vecchia F.M. Bennett S.C. (2007) - A second specimen of the pterosaur Friulano di Storia Naturale, Udine, 319 pp. Anurognathus ammoni. Paläontologische Zeitschrift, 81(4): 376- Dalla Vecchia F.M. (2018) - Comments on Triassic pterosaurs 398. with a commentary on the “ontogenetic stages” of Kell- Bennett S.C. (2014) - A new specimen of the pterosaur Scaphog- ner (2015) and the validity of Bergamodactylus wildi. Rivista nathus crassirostris, with comments on constraint of cer- Italiana di Paleontologia e Stratigrafia, 124(2): 317-341. vical vertebrae number in pterosaurs. Neues Jahrbuch für Dalla Vecchia F. M. (2019) - Seazzadactylus venieri gen. et sp. Geologie und Paläontologie, Abhandlungen, 271(3): 327-348. nov., a new pterosaur (Diapsida: Pterosauria) from the Britt B.B., Dalla Vecchia F.M., Chure D.J., Engelmann G.F., Upper Triassic (Norian) of northeastern Italy. PeerJ, 7: Whiting M.F. & Scheetz R.D. (2018) - Caelestiventus e7363. hanseni gen. et sp. nov. extends the desert-dwelling pter- Dalla Vecchia F.M., Riera V., Oms O., Dinàres-Turell J., Gaete osaur record back 65 million years. Nature Ecology and R., Galobart À. (2013) - The last pterosaurs: First re- Evolution, 2(9): 1386-1392. cord from the uppermost Maastrichtian of the Tremp Cheng X., Wang X., Jiang S. & Kellner A.W.A. (2012) - A new Syncline (Northern Spain). Acta Geologica Sinica, 87(5): scaphognathid pterosaur from western Liaoning, China. 1198-1227. Historical Biology, 24(1): 101-111. Döderlein L. (1923) - Anurognathus ammoni, ein neuer Flug- Clark J.M., Montellano M., Hopson J.A., Hernandez R. & Fa- saurier. Sitzungsberichte der Bayerischen Akademie der Wis- stowsky D.E. (1994) - An Early or Middle Jurassic tet- senschaften, Mathematisch-Naturwissenschaftlichen, 1923: 117- rapod assemblage from the La Boca Formation, North- 164. eastern Mexico. In: Fraser N.C. & Sues H.-D. (eds.) - In Ezcurra M.D., Nesbitt S.J., Bronzati M., Dalla Vecchia F.M., the shadows of the dinosaurs - early Mesozoic tetrapods. Agnolin F.L., Benson R.B.J., Brissón Egli F., Cabreira Cambridge: Cambridge University Press, pp. 295-302. S.F., Evers S.W., Gentil A.R., Irmis R.B., Martinelli A.G., Clark J.M., Hopson J.A., Hernandez R., Fastowsky D.E. & Novas F.E., da Silva L.R., Smith N.D., Stocker M.R., Montellano M. (1998) - Foot posture in a primitive pter- Turner A. H. & Langer M. C. (2020) - Enigmatic dino- osaur. Nature, 391: 886-889. saur precursors bridge the gap to the origin of Pterosau- Codorniú L.S. (2005) - Morfología caudal de Pterodaustro ria. Nature, 588 (7838): 445-449. guinazui (Pterosauria: Ctenochasmatidae) del Cretácico Gao K.Q., Li Q., Wei M., Pak H.& Pak I. (2009) - Early Creta- de Argentina. Ameghiniana, 42(2): 505-509. ceous birds and pterosaurs from the Sinuiju Series, and Colbert E.H. (1969) - A Jurassic pterosaur from Cuba. Ameri- geographic extension of the Jehol Biota into the Kore- can Museum Novitates, 2370: 1-26. an Peninsula. Journal of the Paleontological Society of Korea, Costa F., Alifanov V., Dalla Vecchia F.M. & Kellner A.W.A. 25(1): 57-61. (2013) - On the presence of an elongated tail in an un- Goloboff P. & Catalano S. (2016) - TNT version 1.5, including described specimen of Batrachognathus volans (Pterosau- a full implementation of phylogenetic morphometrics. ria: Anurognathidae: Batrachognathinae). In: Sayão J.M., Cladistics, 32(3): 221-238. Costa F., Bantim R.A.M., Kellner A.W.A. (Eds) - Short Hone D.W.E. (2020) - A review of the taxonomy and palae- communications, Rio Ptero 2013 - International Sym- oecology of the Anurognathidae (Reptilia, Pterosauria). posium on Pterosaurs. Rio de Janeiro: 54-56. Universi- Acta Geologica Sinica, 94(5): 1676-1692. dade Federal do Rio de Janeiro, Museu Nacional, Rio Jensen J.A. & Padian K. (1989) - Small Pterosaurs and Di- de Janeiro. nosaurs from the Uncompahgre Fauna (Brushy Basin Czerkas S.A. & Ji Q. (2002) - A new rhamphorhynchoid with Member, Morrison Formation: ?Tithonian), Late Juras- a headcrest and complex integumentary structures. The sic, Western Colorado. Journal of Paleontology, 63: 364-373. Dinosaur Museum Journal, 1: 16-41. Ji S. & Ji Q. (1998) - A new fossil pterosaur (Rhamphorhyn- Dalla Vecchia F.M. (2002) - A caudal segment of a Late Trias- choidea) from Liaoning. Jiangsu Geology, 22(4):199-206. sic pterosaur (Diapsida, Pterosauria) from Northeastern [in Chinese] Italy. Gortania, 23 (2001): 31-58. Ji Q. & Yuan C.X. (2002) - Discovery of two kinds of proto- Dalla Vecchia F.M. (2009a) - Anatomy and systematics of the feathered pterosaurs in the Mesozoic Daohugou Biota pterosaur Carniadactylus gen. n. rosenfeldi (Dalla Vecchia, in the Ningcheng region and its stratigraphic and biolog- 1995). Rivista Italiana di Paleontologia e Stratigrafia, 115: ic significances. Geological Review, 48(2): 221-224. 159-188. Jiang S., Wang, Cheng X., Rodrigues Costa F., Huang J., Kell- Dalla Vecchia F.M. (2009b) - Austriadactylus (Diapsida, Ptero- ner A.W.A. (2015) - Short note on an anurognathid pter- sauria) from the Norian (Upper Triassic) of northeast- osaur with a long tail from the Upper Jurassic of China. ern Italy. Rivista Italiana di Paleontologia e Stratigrafia, 115: Historical Biology, 27(6): 718-722. 291-304. Kellner A.W.A. (2003) - Pterosaur phylogeny and comments Dalla Vecchia F.M. (2013) - Triassic pterosaurs. In: Nesbitt on the evolutionary history of the group. Geological Socie- S.J., Desojo J.B., Irmis R.B. (Eds) -Anatomy, Phylogeny ty, London, Special Publications, 217: 105-137. and Palaeobiology of Early Archosaurs and their Kin. Kellner A.W.A., Wang X., Tischlinger H., Campos D.A., Hone Geological Society London, Special Publications, 379: 119-155. D.W.E. & Meng X. (2009) - The soft tissue of Jeholopterus Dalla Vecchia F.M. (2014) - Gli pterosauri triassici. Pubblicazi- (Pterosauria, Anurognathidae, Batrachognathidae) and one del Museo Friulano di Storia Naturale n. 54, Museo the structure of the pterosaur wing membrane. Proceed- Orbitoantorbital fenestra in the anurognathids 41 ings of the Royal Society, Series B, 277: 321-329. Sangster S. (2003) - The anatomy, functional morphology and Kuhn O. (1967) - Die fossile Wirbeltierklasse Pterosauria. Oe- systematics of Dimorphodon macronyx (Diapsida: Ptero- ben, Krailling bei München, 52 pp. sauria). PhD thesis, University of Cambridge, Cam- Lü J. (2009) - A new non-pterodactyloid pterosaur from Qin- bridge (UK). glong County, Hebei Province of China. Acta Geologica Sharov A.G. (1971) - New flying reptiles from the Mesozoic Sinica, 83: 189-199. of Kazakhstan and Kirghizia. Trudy Paleontologischeskogo Lü J. & Hone D.W. (2012) - A new Chinese anurognathid Instituta Akademii nauk SSSR, 130: 104-113. [In Russian] pterosaur and the evolution of Pterosaurian tail lengths. Swofford D.L. (2002) - PAUP 4.0b 10. Sinauer Associates, Acta Geologica Sinica, English Edition, 86(6): 1317-1325. Sunderland, Massachusetts. Lü J. & Ji Q. (2006) - Preliminary results of a phylogenetic Tischlinger H. & Frey E. (2013) - Ein neuer Pterosaurier analysis of the pterosaurs from Western Liaoning and mit Mosaikmerkmalen basaler und pterodactyloider surrounding areas. Journal of the Paleontological Society of Pterosauria aus dem Ober-Kimmeridgium von Painten Korea, 22: 239-261. (Oberpfalz, Deutschland). Archaeopteryx 31: 1-13. Lü J., Ji S., Yuan C. & Ji Q. (2006) - Pterosaurs from China. Geo- Unwin D.M. (2003) - On the phylogeny and evolutionary his- logical Publishing House, Beijing, 147 pp. tory of pterosaurs. Geological Society, London, Special Publi- Lü J., Xu L., Chang H. & Zhang X. (2011) - A new darwinop- cations, 217: 139-190. terid pterosaur from the Middle Jurassic of Western Li- Unwin D.M. (2006) - The pterosaurs from deep time. Pi Press, aoning, Northeastern China and its ecological implica- New York, 347 pp. tions. Acta Geologica Sinica, 85: 507-514. Unwin D.M. & Bakhurina N.N. (2000) - Pterosaurs from Rus- Lü J., Unwin D.M., Jin X., Liu Y. & Ji Q. (2010a) - Evidence sia, Middle Asia and Mongolia. In: Benton M.J., Shishkin for modular evolution in a long-tailed pterosaur with a M.A., Unwin D.M. & Kurochkin E.N. (eds.), The age pterodactyloid skull. Proceedings of the Royal Society B, 7: of dinosaurs in Russia and Mongolia. Cambridge: Cam- 383-389. bridge University; pp. 420-433. Lü J., Fucha X. & Cheng J. (2010b) - A new scaphognathine Unwin D. M., Lü J., Bakhurina N.N. (2000) - On the system- pterosaur from the Middle Jurassic of western Liaoning, atic and stratigraphic significance of pterosaurs from China. Acta Geoscientica Sinica, 31(2): 263-266. the Lower Cretaceous Yixian Formation (Jehol Group) Lü J., Unwin D.M., Bo Z., Chunling G. & Caizhi S. (2012) of Liaoning, China. Mitteilungen aus dem Museum für - A new rhamphorhynchid (Pterosauria: Rhamphorhy- Naturkunde Berlin, Geowissenschaftliche Reihe, 3(1): 181-206. nchidae) from the Middle/Upper Jurassic of Qinglong, Vidović S.U. & Martill D.M. (2018). The taxonomy and phy- Hebei Province, China. Zootaxa, 3158: 1-19. logeny of Diopecephalus kochi (Wagner,1837) and ‘German- Lü J., Meng Q., Wang B., Liu D., Shen C. & Zhang Y. (2018) odactylus rhamphastinus’ (Wagner, 1851). Geological Society, - Short note on a new anurognathid pterosaur with evi- London, Special Publications, 455(1):125-147. dence of perching behaviour from Jianchang of Liaon- Wang X., Zhou Z., Zhang F. & Xu X. (2002) - A nearly com- ing Province, China. Geological Society, London, Special Pub- pletely articulated rhamphorhynchoid pterosaur with ex- lications, 455: 95-104. ceptionally well preserved wing membranes and ‘hairs’ Lü J., Pu H., Xu L., Wei X., Chang H. & Kundràt M. (2015) - from Inner Mongolia, northeast China. Chinese Science A new rhamphorhynchid pterosaur (Pterosauria) from Bulletin, 47(3): 226-230. Jurassic deposits of Liaoning Province, China. Zootaxa, Wang X., Kellner A.W.A., Jiang S.X. & Meng X. (2009) - An 3911 (1): 119-129. unusual long tailed pterosaur with elongated neck from Nesbitt S.J. (2011) - The early evolution of archosaurs: rela- western Liaoning of China. Anais da Academia Brasileira tionships and the origin of major clades. Bulletin of the de Ciências, 81(4): 793-812. American Museum of Natural History, 352: 1-292. Wang X., Kellner A.W.A., Jiang S., Cheng X., Meng X. & Nopcsa F. (1928) - The genera of reptiles. Palaeobiologica, 1: Rodrigues T. (2010) - New long-tailed pterosaurs (Wu- 163-188. kongopteridae) from western Liaoning, China. Anais da Ősi A., Prondvai E., Frey E. & Pohl B. (2010) - New Interpre- Academia Brasileira de Ciências, 82(4): 1045-1062. tation of the Palate of Pterosaurs. The Anatomical Record, Wang X., Rodrigues T., Jiang S.X., Cheng X. & Kellner A.W.A. 293: 243-258. (2014) - An Early Cretaceous pterosaur with an unusual Padian K. (2008a) - The Early Jurassic pterosaur Dorygnathus mandibular crest from China and a potential novel feed- banthensis (Theodori, 1830). Special Papers in Palaeontology, ing strategy. Scientific Reports, 4: 6329. 80: 1-64. Wang X., Jiang S.X., Zhang J., Cheng X., Yu X., Li Y., Wei J. Padian K. (2008b) - The Early Jurassic pterosaur Campylogna- & Wang X. (2017) - New evidence from China for the thoides Strand, 1928. Special Papers in Palaeontology, 80: 65- nature of the pterosaur evolutionary transition. Scientific 107. Reports, 7: 42763, 9 pp. Romer A.S. (1956) - Osteology of Reptiles. 3rd edition (1976). Wei X., Pêgas R.V., Shen C., Guo Y., Ma W., Sun D. & Zhou X. University Press, Chicago & London, xxi + 772 pp. (2021) - Sinomacrops bondei, a new anurognathid pterosaur Ryabinin A.N. (1948) - Remarks on a flying reptile from the from the Jurassic of China and comments on the group. Jurassic of Karatau. Trudy Paleontologischeskogo Instituta PeerJ, 9: e11161. Akademii nauk SSSR, 15(1): 86-93. [in Russian] Wellnhofer P. (1970) - Die Pterodactyloidea (Pterosauria) 42 Dalla Vecchia F.M. der Oberjura-Plattenkalke Süddeutschlands. Bayerische Witton M.P. (2013) - Pterosaurs: natural history, evolution, Akademie der Wissenschaften Mathematisch-Wissenschaftlichen anatomy. Princeton University Press, Princeton ad Ox- Klasse, Abhandlungen, 141: 1-33. ford, 291 pp. Wellnhofer P. (1975a) - Die Rhamphorhynchoidea (Pterosau- Wu W.H., Zhou C.F. & Andres B. (2017) - The toothless ptero- ria) der Oberjura-Plattenkalke Süddeutschlands. Teil I: saur Jidapterus edentus (Pterodactyloidea: Azhdarchoidea) Allgemaine Skelettmorphologie. Palaeontographica, A 148: from the Early Cretaceous Jehol Biota and its paleoeco- 1-33. logical implications. PLoS ONE, 12(9): e0185486. Wellnhofer P. (1975b) - Die Rhamphorhynchoidea (Pterosau- Yang Z., Jiang B., McNamara,M.E., Kearns S.L., Pittman M., ria) der Oberjura-Plattenkalke Süddeutschlands. Teil II: Kaye T.G., Orr P.J., Xu X. & Benton M.J. (2019) - Ptero- Systematische Beschereibung. Palaeontographica, A 148: saur integumentary structures with complex feather-like 132-186. branching. Nature ecology & evolution, 3: 24-30. Wellnhofer P. (1978) - Pterosauria. In: Wellnhofer P. (Ed.) - Zhou C.-F. (2014) - Cranial morphology of a Scaphognathus-like Handbuch der Paläoherpetologie, Part 19. Gustav Fish- pterosaur, Jianchangnathus robustus, based on a new fossil er, Stuttgart, 82 pp. from the Tiaojishan Formation of western Liaoning, Wellnhofer P. (1991) - The Illustrated Encyclopedia of Ptero- China, Journal of Vertebrate Paleontology, 34(3): 597-605. sauria. Salamander Book, London, 192 pp. Zhou C.-F. & Schoch R.R. (2011) - New material of the Wild R. (1979) - Die Flugsaurier (Reptilia, Pterosauria) aus der non-pterodactyloid pterosaur Changchengopterus pani Lü, Oberen Trias von Cene bei Bergamo, Italien. Bollettino 2009 from the Late Jurassic Tiaojishan Formation of della Società Paleontologica Italiana, 17: 176-256. western Liaoning. Neues Jahrbuch für Geologie und Paläontol- ogie, Abhandlungen, 260: 265-275.