Anhanguera taxonomy revisited: is our understanding of Santana Group pterosaur diversity biased by poor biological and stratigraphic control?

Anhanguera sp.

Locality and horizon. Romualdo Formation, Araripe Basin, Albian, northeastern Brazil. According to Wellnhofer (1991), the specimen comes from Jardim municipality in the state of Ceará (previously Barra do Jardim), but its exact locality is undetermined.

Anatomical description. The skull of AMNH 22555 is nearly complete and best preserved in right lateral aspect (Figs. 4 and 5). Even so, rostral elements anterior to the nasoantorbital fenestrae are crushed and laterally compressed in this view. Posterior skull bones are broken and disarticulated in the left lateral view, in which the absence of bones such as the left jugal and lacrimal obliterates the edges of skull openings. The palate anterior to the choanae is well preserved, whereas posterior palatal bones are mostly absent. Parts of the right pterygoid lie inside the nasoantorbital opening in lateral view. Despite the fact that the alveolar margin of the maxillae is intact anteriorly, with the presence of some in situ teeth (mostly broken) and empty alveoli, the ventral margins of both the left and right maxillae are eroded and incomplete posteriorly, preventing an accurate estimation of the total number of tooth positions. The posterior skull roof is almost intact, with a slight lateral displacement of the frontoparietals. Above the nasoantorbital openings, the outer bone layer of the dorsal margin of the fused premaxillae is eroded. The dorsal limits of the premaxillae are badly crushed throughout the anterior half of the skull, preventing the reconstruction of the sagittal crest anatomy. In occipital view, only the broad supraoccipital plate and right opisthotic are fairly well preserved.

In general, the skull bones are disarticulated and, sometimes, displaced from their original positions. The premaxillae and maxillae, as well as the frontals and parietals, are tightly fused with each other, displaying the ordinary condition for pterodactyloids. Some postcranial bones, known to fuse in mature individuals, show the unfused condition in AMNH 22555, indicating that this specimen is osteologically immature (Wellnhofer, 1991; Bennett, 1993). Those elements include separate scapulae and coracoids, as well as proximal and distal carpals (Figs. 2H–2K). The first five dorsal vertebrae show very thick neural spines and prezygapophyses fused with the postzygapophyses of the adjacent vertebra, indicating that a notarium was present in mature individuals of this species (Fig. 2B).

Premaxilla. The fused premaxillae comprise most of the skull roof, with their posterior ends dorsal to the orbits, where they contact the frontoparietals posteriorly. Although the left premaxilla is considerably well preserved throughout its whole extension, the right element is badly crushed anteriorly to the nasoantorbital fenestra. Sutures between the premaxillae and maxillae can only be observed close to the nasoantorbital fenestrae, especially on the left side of the skull (where this region is best preserved). Anteriorly, the ventral limits of the premaxillae are not clear, and the number of tooth positions associated with these bones cannot be inferred. The dorsal surface of the premaxillae is broken in the region anterior to the nasoantorbital fenestrae, making it difficult to determine if a sagittal crest was present. However, this broken dorsal border extends above the projection of the surface dorsal to the nasoantorbital openings, which may indicate that the crest was present. It is probable that the premaxillae also composed the anterior part of the palate, where the bone is strongly pierced by small foramina. However, due to bone fusion, it is impossible to determine the exact contribution of the premaxillae to the palatal surface. There is a discrete anterior expansion of the skull, with the rostrum being about 1.5–2 mm wider at the level of the 4th tooth sockets than at the 3rd and 5th alveoli. This is more reminiscent of the slight expansion seen in Tropeognathus mesembrinus, but at this point it cannot be ruled out that the expansion could grow larger with maturity.

Maxilla. Bordered dorsally by the premaxillae, the maxillae form the anterior and part of the ventral margins of the nasoantorbital fenestrae. Because the suture lines between the maxillae and premaxillae are located at the anterodorsal border of the nasoantorbital fenestrae, the maxillae also make a small contribution to the dorsal margin of these openings. Ventrally, the palatal plates of the maxillae fuse together (see Ősi et al., 2010; Pinheiro & Schultz, 2012), forming a well-developed palatal ridge that ends about 50 mm before the anterior limits of the choanae. The dental margins of the maxillae form strong rims, and some of the rostral teeth (especially the 7th to 10th tooth pairs) are surrounded at their bases by robust bony collars, generally punctured by foramina on their lingual side. Because the jugal processes of both maxillae are broken, the posterior limits of these bones cannot be determined. Anterior to the 9th tooth pair, the ventral margins of the maxillae gently curve upwards, and the anteriormost teeth are inserted at level with the ventral margins of the orbits.

Nasal. Together with the lacrimals, the nasals form the posterodorsal margins of the nasoantorbital fenestrae. The right nasal is better preserved than the left one, and shows an irregular shape, with acute anterior and posterior extensions. The dorsal margin is straight and contacts the premaxillae. The nasals have lateral longitudinal ridges, probably indicating the contact area with the lacrimals (in AMNH 22555, these bones are slightly displaced). The nasals have concave posterior margins, fitting the convex prefrontals and supraorbitals. The ventral surfaces of the acute anterior processes of the nasals are perforated by well-developed foramina. The nasoantorbital openings are completely filled with carbonaceous matrix and, thus, the medial contact between the left and right nasals, as well as the ventral nasal process, are obscured.

Prefrontal. Only the right prefrontal is preserved. Dorsally, this bone makes contact with the nasal and the supraorbital, whereas ventrally it shows a rectilinear suture with the lacrimal. The prefrontal contributes to part of the anterodorsal margin of the orbit.

Supraorbital. Both supraorbitals are preserved. These bones are roughly triangular in dorsal aspect and compose part of the skull roof above the orbits. The contact between the supraorbitals and frontoparietals is marked by grooves, which are deeper at their posterior limits. The supraorbitals are also partially covered by the posterior extension of the premaxillae.

Frontoparietal. There is no visible distinction between the frontals and parietals, but a clear suture line divides the left and right elements of these bones. The frontoparietals form almost the entire skull roof above the orbits and the upper temporal fenestrae, being overlaid anteriorly by the slender posterior extension of the premaxillae that projects between the left and right frontoparietals. Above the upper temporal fenestrae, the dorsal margin of the frontoparietals forms a short crest that probably provided a greater area of origin for the musculus adductor mandibulae externus.

Jugal. Only the right jugal is preserved. This is a robust element, mostly composed of three strong processes that contribute to the boundaries of several skull openings. The maxillary process of the jugal extends anteriorly, forming part of the posteroventral margin of the nasoantorbital fenestra as well as it contributes to the lateral margin of the palatal subtemporal fenestra. This process is broken in the preserved jugal of AMNH 22555, preventing an estimation of how far anteriorly the contact with the maxillae was located. The lacrimal process of the jugal is directed dorsally, with a slight anterior inclination, and forms part of the anterior margin of the orbit, as well as part of the posterior margin of the nasoantorbital fenestra. The spot where this process connects with the main corpus of the jugal is depressed, forming a distinct lateral shelf, so that the whole process is medially displaced with respect to the remainder of the bone. The contact with the lacrimal occurs at about one fourth of the total height of the orbit. The most developed jugal process is the posterior, postorbital, one. This bony extension is very thick anteriorly, but becomes narrower throughout its posterodorsal end, where it contacts the postorbital via an overlapping joint. The postorbital process of the jugal composes most of the posterior edge of the orbit, and the whole anterior border of the lower temporal fenestra.

Postorbital. Both postorbitals are preserved; the left one is completely displaced from its original position and the right one shows a slight medial displacement. These bones have a roughly triangular outline and occupy a central position on the temporal region of the skull. The postorbitals make contact dorsally with the frontoparietals, anteriorly with the posterior processes of the jugals and posteriorly with the squamosals. The edges of these bones contribute to the margins of both the upper and lower temporal fenestrae, and also have a small participation in the posterior borders of the orbits.

Lacrimal. In AMNH 22555, only the right lacrimal is preserved. This bone is triangular in shape, making contact with the prefrontal and the nasal dorsally and overlying the lacrimal process of the jugal ventrally. The posterior edge of the lacrimal bears a well-developed, lateromedially broad process directed inside the orbit. The lacrimal is pierced by a vast foramen for the exit of the naso-lacrimal duct, which occupies most of the main corpus of this bone.

Squamosal. The squamosal is a curved bone, with its concavity directed anteriorly, where this element comprises most of the posterior border of the lower temporal fenestra. Dorsally, the squamosal contacts the postorbital and frontoparietals. Between these bones there is a smaller concavity that bounds the ventral margin of the upper temporal fenestra. The squamosal ends ventrally with two acute processes. The anterior one sutures with the slim quadratojugal, whereas the posterior one runs parallel to the quadrate and is probably the origin site of the musculus depressor mandibulae. The posterior, convex edge of the squamosal makes contact with the opisthotic.

Quadratojugal. This slender bone makes contact with the main corpus of the jugal anteriorly and with one of the ventral processes of the squamosal posteriorly, delimiting ventrally the lower temporal fenestra.

Quadrate. Only the right quadrate is completely preserved. This bone contacts the squamosal, quadratojugal and part of the jugal. The anteroventral end of the quadrate expands to form the helical articular surface with the lower jaw. The quadrate shaft runs medially, parallel to the ventral extension of the squamosal. The inclination of the quadrate with respect to the ventral margin of the maxilla is about 145 degrees.

Supraoccipital. The supraoccipital is a broad plate that forms a large portion of the occiput. Above the dorsal margin of the foramen magnum, this bone develops a low sagittal crest, probably linked to the origin of the musculus rectus capitis. Lateral to the crest, the supraoccipital is pierced by two large pneumatic foramina. The dorsal border of the right posttemporal fenestra is preserved, showing that this opening was inclined downwards (Fig. 5).

Opisthotic. In AMNH 22555, both the right and left opisthotics are broken and displaced from their original positions. Although the right element is better preserved, little anatomical information can be drawn from this bone. It can be observed that the opisthotics were configured as wide plates that occupied a considerable portion of the occiput.

Palatine. The structures traditionally regarded, in most pterosaurs, as the palatines were recently reinterpreted as a secondary surface formed by ventral plates of the maxillae (see Ősi et al., 2010; Pinheiro & Schultz, 2012). The high degree of synostosis, common in Pterodactyloidea, makes the individualization of palatal elements difficult. In anhanguerids, the palatines probably bordered the suborbital fenestrae medially, the right element being partially preserved in AMNH 22555 (Fig. 4C; Pinheiro & Schultz, 2012: Figs. 4C and 4D).

Pterygoid. Although most of the posterior palatal bones were lost, part of the left pterygoid lies in dorsal view inside the nasoantorbital fenestra. This bone shows a very long and acute rostral process connected to a concave surface, which is followed posteriorly by a transversal ridge. This ridge can be interpreted as part of a vestigial ectopterygoid, already reported for other specimens of Anhanguera (Pinheiro & Schultz, 2012). In close association, there is a flat, triangular bone of uncertain affinities. It is possible that it represents the posterior extension of the pterygoid, which would contact the basipterygoid caudally.

Vomers. The fused vomers form a slim element that partially divides the choanae medially. There is no sign of sutures between the two vomers or between them and other elements of the palate.

Dentition. Only the dentition pattern of the upper jaw of AMNH 22555 can be assessed, and it is reminiscent of that seen in other species of Anhanguera. The 1st pair of teeth is located at the tip the rostrum, slightly higher than the 2nd pair, facing anteriorly, as is usual in anhanguerians (Rodrigues & Kellner, 2013). The alveoli grow in width until the 3rd pair. As is usual in the genus Anhanguera, the 4th and 7th pairs of alveoli are larger than the 5th and 6th. From the 8th onwards, the alveoli tend to gradually decrease in width. The distances between the alveoli increase gradually, but are most notably larger from between the 7th and 8th alveoli onwards. As noted above, the maxillary margin is not well preserved and most posteriormost alveoli cannot be assessed, but the dentition would continue until at least the beginning of the nasoantorbital fenestra. Some teeth are preserved, showing a curved and pointed shape and longitudinal ridges where the enamel is present, as typical of anhanguerids (Rodrigues & Kellner, 2010).

On the diagnosis of Anhanguera

Kellner (2003) listed synapomorphies of the genus Anhanguera as (1) the presence of an elongate and medially placed nasal process, (2) a foramen on the nasal process, (3) a characteristic size difference in the rostral teeth (in which the 5th and 6th tooth pairs are smaller than the 4th and 7th ones); (4) scapulae length at most 80% of that of the coracoids, (5) a coracoidal articulation surface with the sternum oval and with a posterior expansion, and (6) a pneumatic foramen on the proximal dorsal surface of the humeri. However, more recently described specimens challenge some of these features and show that they are more widespread among dsungaripteroid pterosaurs. Characters (1) and (2) are present on Ludodactylus sibbicki from the Crato Formation (Frey, Martill & Buchy, 2003), and characters (4), (5) and (6) are also found in Brasileodactylus sp. (SNSB-BSPG 1991 I 27; Veldmeijer, Meijer & Signore, 2009) and in Istiodactylus (Hooley, 1913; Andres & Ji, 2006). Therefore, from these, only character (3) would be unambiguously synapomorphic for Anhanguera.

Naturally, these are characters used in a cladistic sense, but others have also been proposed as diagnostic of the genus. While comparing Anhanguera and Coloborhynchus, Fastnacht (2001) stated that Anhanguera possesses (1) a premaxillary crest beginning more posteriorly instead of at the anterior tip of the rostrum, (2) a premaxillary crest lower than in Coloborhynchus with its height about one third of its length, (3) a thin crest, (4) the anterior end of the rostrum inclined at an angle of about 45 degrees, and (5) the absence of a spoon-shaped distal expansion of the rostrum. From these, our analyses demonstrate that characters (1) and (2) could be attributed to ontogenetic development in the genus Anhanguera. Character (5) is a misinterpretation since the type species, Anhanguera blittersdorffi, has a distal expansion with this morphology (see Rodrigues & Kellner, 2008). Characters (3) and (4), although useful to distinguish Anhanguera from Coloborhynchus, are also present in Liaoningopterus and Caulkicephalus (Wang & Zhou, 2003; Steel et al., 2005; Rodrigues et al., 2015) and therefore are more widespread within anhanguerids. A very similar set of characters was also discussed by Veldmeijer (2003). This author suggested that AMNH 22555 is a juvenile Coloborhynchus. However, some of the characters used by him to separate Anhanguera and Coloborhynchus, such as the position of the premaxillary crest, are also listed as being possibly explained by ontogenetic variation, a view that is supported by our results. Veldmeijer (2003) also suggested that features present at the posterior part of the skull of AMNH 22555 are more similar to Anhanguera spielbergi (regarded by him as belonging to the genus Coloborhynchus) than to the holotype of “Anhanguera santanae”. However, the diagnostic value of these minor differences is dubious. Therefore, it seems that Anhanguera remains diagnosed by a single unambiguous character state, the 5th and 6th tooth pairs being smaller than the 4th and 7th ones, and by combinations of characters.

Here we suggest the following revised diagnosis for Anhanguera, which incorporates the ontogenetic changes discussed above: anhanguerid pterosaurs with premaxillary and dentary median crests at least in advanced ontogenetic stages; premaxillary crest thin; premaxillary crest largely asymmetric; premaxillary crest begins near but not at the tip of the skull; premaxillary crest not confined to the anteriormost tip of the skull; premaxillary crest grows allometrically in height and length during ontogeny; 5th and 6th upper dental alveoli smaller than the 4th and 7th ones; parietal crest blade-like and thin; palatal ridge modest in depth.

AMNH 22555 cannot be confidently referred to what is known as “Anhanguera santanae”

When first described by Wellnhofer (1991), AMNH 22555 was referred to “Anhanguera santanae”, a pterodactyloid pterosaur described a few years before by the same author and from the same formation (Wellnhofer, 1985). The assignment of AMNH 22555 to “A. santanae” (then regarded as “Araripesaurus”; see Introduction) was made mainly on the basis that both specimens share the same number of bones in the carpal series, besides possessing similar sized skulls, even though the position of their premaxillary crests differed (Wellnhofer, 1991). A close examination of the “A. santanae” holotype (SNSB-BSPG 1982 I 90) and comparison to other skulls now known, however, has revealed to us that AMNH 22555 cannot be confidently referred to this species more than to any other proposed species of Anhanguera.

Although AMNH 22555 is indeed similar to the “Anhanguera santanae” holotype in size and overall skull morphology, the two specimens differ in a series of features (Fig. 5). First of all, the frontoparietals of “A. santanae” are relatively narrower and project posterodorsally as a thick frontoparietal crest. On the other hand, the frontoparietals of AMNH 22555 are broader and form a much more delicate crest, which is mostly posteriorly extended. The two specimens also differ in the morphology of the jugal: the lacrimal process of this bone is much broader in A. santanae than in AMNH 22555.

Differences between AMNH 22555 and “A. santanae” also extend to the occipital and palatal regions. In occipital view, it is notable that the supraoccipital crest is much more conspicuous in “A. santanae” than in AMNH 22555. Also, although the occiput of AMNH 22555 is not well preserved, the dorsal margin of the posttemporal fenestra is well marked and reveals that this opening was probably directed downwards, unlike the condition observed in the “A. santanae” holotype. As a consequence of the poor preservation, however, this characteristic must be regarded with caution. In palatal view, it is remarkable that in AMNH 22555 the fusion of the palatal plates of the maxillae develops into a strong palatal ridge (although not as deep as observed in Tropeognathus) that is followed posteriorly by a slight convexity of the palatal occlusal surface. “A. santanae” also bears a palatal ridge, but this structure is much less pronounced and extends posteriorly to a region closer to the choanae than that seen in AMNH 22555. In addition, the choanal morphology is also different between the specimens, with those of AMNH 22555 being distinctly rounder and more lateromedially expanded.

In spite of these remarkable differences between AMNH 22555 and the “Anhanguera santanae” holotype (SNSB-BSPG 1982 I 90), none of the characters listed above have had their distributions well mapped for Anhanguera, and may fall within the range of intraspecific variation for this genus. In addition, it is noteworthy that allegedly diagnostic features of Anhanguera nominal species are, in most cases, subtle and poorly defined, especially those which are related to the presence and morphology of the premaxillary crest. As discussed, the premaxillary crest shows significant allometric growth within Anhanguera-like pterodactyloids, demonstrating that this structure is at least partially body size-dependent and therefore has limited use for taxonomic purposes. Bearing this in mind, we reassess here the significance of anatomical features of the premaxillary crest traditionally thought to support Anhanguera species, and elucidate the impact of this on the taxonomy of this genus.

On the validity of “Anhanguera santanae” and other species of Anhanguera

A reappraisal of the purportedly diagnostic features of the individual Anhanguera species revealed that most, if not all, of the characters that are currently used to define the separate species are probably well inside the range of intraspecific variation. Considering this, it is pertinent to inquire about the validity of each one of the species attributed to this genus.

When first described, “Anhanguera santanae” was differentiated from other pterosaurs on the basis of characters that are today known to be widely distributed among other Santana Group ornithocheiroids. A complete discussion of the validity of the diagnostic features originally proposed for “A. santanae” was made by Kellner & Tomida (2000). As a conclusion, these authors stated that the only remaining diagnostic character for this taxon would be the position of the premaxillary crest, well anterior to the nasoantorbital fenestrae. It is noteworthy that the premaxillary crest itself is not preserved on “A. santanae” holotype, and its presence is inferred by the acute dorsal margin of the premaxillae close to the anterior extremity of the specimen, as preserved. One of the specimens analyzed in the present allometric regression, SMNK PAL 1136, presents a premaxillary crest that can be presumed to be positioned as far from the nasoantorbital opening as inferred in the holotype of “A. santanae”. As discussed, premaxillary crest characters are here regarded as unfit for the diagnosis of nominal anhanguerid species, what means that “A. santanae” holotype lacks unambiguous diagnostic features and should be considered as a nomen dubium.

Anhanguera blittersdorffi, the type species of Anhanguera, was first diagnosed by characters that later proved to be diagnostic of more inclusive clades, such as the presence and morphology of the premaxillary and frontoparietal crests and the presence of a distal expansion and of larger teeth at the tip of the rostrum (Campos & Kellner, 1985). Actually, A. blittersdorffi has the standard morphology of Anhanguera and, after the description of other anhanguerids, it became difficult to recognize unique characters for this species. A subsequent revision (Kellner & Tomida, 2000) proposed diagnostic characters of the species as a “lower skull with a proportionally shorter quadrate”. Those characters, however, are subjective and ambiguous, and lack a quantified definition to delimit the state boundaries. Besides the holotype, only one additional specimen has been formally attributed to A. blittersdorffi (Pz-DBAV-UERJ 40) (Kellner & Tomida, 2000), though the latter still lacks an anatomical description. Remarkably, the number of alveoli on A. blittersdorffi upper jaws (52) is higher than in any other proposed Anhanguera species and this might be a more suitable diagnostic character for this taxon.

“Anhanguera araripensis” was described based on a very incomplete skull with associated postcranial bones by Wellnhofer (1985). As was the case for A. blittersdorfii, “A. araripensis” was first diagnosed by characters that later were demonstrated to be more widespread among anhanguerids or dependent on ontogenic status of specimens. After the revision of Kellner & Tomida (2000), only two characters remained as diagnostic for this species: the dorsal margin of the premaxillae being “keel shaped” up to the anterior end of the nasoantorbital fenestrae (a character described as being related to the position of the premaxillary crest, which is not preserved at the holotype), and the presence of small lateral projections on the basioccipital processes of the pterygoids (Kellner, 1991) (Fig. 6D). Based on this character, other specimens have also been referred to this species, such as MN 4735-V (Kellner & Tomida, 2000) and SAO 16494 (Veldmeijer, 2003; Veldmeijer, 2006). We agree that the “keel shaped” dorsal margin of the premaxillae is probably related to the presence and morphology of the premaxillary crest and, for the reasons described above, challenge the taxonomic value of this character. Regarding the lateral projections of the pterygoids inside the subtemporal fenestrae, we consider this character as problematic, because it is probably related to the bone growth between different elements of the adductor musculature that crossed the subtemporal openings. Also, these projections are exceptionally delicate and were probably abraded on not so well preserved skulls. Remarkably, specimens such as the holotype of Tropeognathus mesembrinus (SNSB-BSPG 1987 I 46) and A. blittersdorffi (MN 4805-V) have very discrete bulges at this same location (Fig. 6E). Thus, we here regard the holotype of “A. araripensis” as nondiagnostic and, for this reason, “Anhanguera araripensis” should also be considered as a nomen dubium.

“Anhanguera robustus” was originally referred to the genus Tropeognathus by Wellnhofer (1987) and later assigned to Anhanguera (Kellner & Campos, 1988). This taxon was originally diagnosed by the presence of a well-developed dentary crest, with a straight anterior margin; and by a spoon-like anterior expansion of the dentaries and long anterior teeth. As has already been observed by Kellner & Tomida (2000), strong anterior teeth associated to a lateral expansion of the dentaries are considered to be widespread among anhanguerids. The other supposedly diagnostic characters are related to the dentary sagittal crest and are probably associated to the apparently advanced ontogenetic stage of the specimen (SNSB-BSPG 1987 I 47). Thus, we also consider “Anhanguera robustus” as a nomen dubium.

Despite its large body size, the holotype of Anhanguera piscator presents clear evidence of an early ontogenetic stage, which partially explains the presence of the low premaxillary crest that was regarded by Kellner & Tomida (2000) as diagnostic for this species. Our analysis demonstrates that premaxillary crest height in this species cannot be explained by allometric growth alone, but nonetheless we consider this character alone to be inappropriate for the diagnosis of anhanguerids. Kellner & Tomida (2000) indicated another cranial character as diagnostic for this taxon: a “basisphenoid constricted in the middle part” (Kellner & Tomida, 2000, p. 7). This feature cannot be accessed in SNSB-BSPG 1982 I 89 (“Anhanguera araripensis”) or SNSB-BSPG 1987 I 47 (“Anhanguera robustus”). Although this character still lacks an unambiguous morphometric definition, basisphenoid morphology in A. piscator holotype is indeed different from what is observed in Anhanguera blittersdorffi and SNSB-BSPG 1982 I 90 (“Anhanguera santanae”), resembling the condition of Anhanguera spielbergi. Other proposed diagnostic features of A. piscator are associated to the postcranial skeleton, which is poorly preserved or absent in most other Anhanguera holotypes. A. piscator is here retained as a valid taxon, at least until more information about the distribution of these postcranial characters and the basisphenoid morphology becomes clearer within Anhangueridae.

Veldmeijer (2003) considered Anhanguera spielbergi to be a representative of Coloborhynchus, including in the diagnosis of this species an “ill-defined, almost absent (…) palatinal ridge and corresponding mandibular groove; mandibular groove not extending onto spoon-shaped expansion; slight, almost absent, ventrolaterally extending tooth-bearing maxillae; large premaxillary sagittal crest, in ratio length-total length skull, which extends dorsally from the anterior aspect until the anterior border of the nasoantorbital fenestra; strongly medial bended rami; sternum with rounded triangular posterior plate of which the length is as long as the width” (Veldmeijer, 2003, p. 43). Although the palatal ridge of A. spielbergi is indeed weaker than that which is observed in other Anhanguera holotypes, it is still not clear how this character is affected by ontogeny, the same also being a potential issue for the mandibular groove morphology. As we have discussed, premaxillary crest morphology is here regarded as inappropriate for species-level diagnoses. Furthermore, a medial bending of mandibular rami cannot be assessed in most of the other holotypes, but is present in other complete anhanguerid mandibles (for instance, “Anhanguera robustus”). Although the intrageneric variation of the remaining characters is still unclear, we regard A. spielbergi as a valid taxon, a taxonomic statement that requires further testing through more comprehensive sampling within the genus.