Anhanguerids comprise an important clade of pterosaurs, mostly known from dozens of three-dimensionally preserved specimens recovered from the Lower Cretaceous Romualdo Formation (northeastern Brazil). They are remarkably diverse in this sedimentary unit, with eight named species, six of them belonging to the genus
We describe in detail the skull of one of the most complete specimens referred to
Geometric morphometric regression of shape on centroid size was highly statistically significant (
Historically, minor differences in crest morphology have been used in the definition of new anhanguerid species. Nowadays, this practice resulted in a considerable difficulty in referring well-preserved skulls into known taxa. When several specimens are analyzed, morphologies previously believed to be disparate are, in fact, separated by a continuum, and are thus better explained as individual or temporal variations. Stratigraphically controlled excavations on the Romualdo Formation have showed evidence for faunal turnover regarding fish communities. It is thus possible that some of the pterosaurs from this unit were not coeval, and might even represent anagenetic morphotypes. Unfortunately, amateur collecting of Romualdo Formation fossils, aimed especially at commerce, resulted in the lack of stratigraphic data of virtually all its pterosaurs and precludes testing of these further hypotheses.
Anhangueridae is a clade of pterosaurs currently known from multiple localities worldwide, including named species from Brazil, the United States, Morocco, China and England (
Although the first descriptions of pterosaurs from the Romualdo Formation date from as early as the 1970s (
Additional anhanguerid specimens, but no newly named species, were subsequently described by
Synopsis of the named species of
Species | Taxonomic status in the present work | Known specimens | Diagnosis |
---|---|---|---|
Type-species | MN 4805-V (holotype) Pz-DBAV UERJ 40 | Large number (52) of alveoli on the upper jaw | |
SNSB-BSPG 1982 I 89 (holotype) | Non diagnostic | ||
SNSB-BSPG 1982 I 90 (holotype) | Non diagnostic | ||
SNSB-BSPG 1987 I 47 (holotype) | Non diagnostic | ||
Valid | NSM-PV 19892 (holotype) | From |
|
Valid | RGM 401 880 (holotype) | From |
Specimen | Taxonomic history | Taxonomic status in the present work |
---|---|---|
AMNH 22555 | Referred to “ |
|
MN 4735-V | Referred to “ |
|
NHMUK R 11978 | None | |
SAO 16494 | Referred to “ |
|
SMNK PAL 1136 | Referred to |
Today, several skulls (both described and undescribed) are hosted in a myriad of publically accessible collections and thus enabling the examination of a larger sample of Romualdo anhanguerids (
Here we reanalyze the skull of the specimen AMNH 22555, originally referred to “
All the specimens up until now assigned to the genus
Levels where pterosaur fossils are found are indicated. Modified from
In order to assess the biological and stratigraphic biases that may have impacted on the taxonomy of
(A) pelvic region in dorsal view; (B) torso in dorsal view; (C, D, E) sixth cervical vertebrae in, respectively, anterior, dorsal and right lateral views; (F, G) right mandibular ramus in, respectively, medial and lateral views; (H) left scapula in dorsal view; (I) left coracoid in lateral view; (J) distal carpals in distal view; (K) proximal carpals in distal view. Scale bars equal to 50 mm. Line drawings of some bones were modified from
In order to assess size-dependent characters within
Two-dimensional coordinates were captured for 17 landmarks using digital photographs of specimens in lateral aspect and the software TPSDig (
Used landmarks are plotted in the skull of
The main goal of our analyses was to detect and describe morphologic variation attributable to the increase of skull size, especially with respect to the premaxillary crest. Although our study is mainly focused on the genus
As the landmark plotting for
Our first analysis, including the whole sample of 12 skulls attributed to
The second analysis, in which
We also analyzed the residual (uncorrelated with size) component of variation for each specimen, in an attempt to identify individual morphological disparity, which is potentially attributable to interspecific variation. At least two specimens indeed show a considerable amount of residual variation of shape, unpredicted by our regression model. Specimen MN 4735-V, attributed by
Pterosauria |
Pterodactyloidea |
Anhangueria |
Anhangueridae |
(A) right lateral, (B) dorsal and (C) palatal views. Abbreviations: ch, choanae; ec, ectopterygoid; fp, frontoparietal; j, jugal; l, lacrimal; m, maxilla; n, nasal; naof, nasoantorbital fenestra; op, opisthotic; pf, prefrontal; pl, palatine; po, postorbital; pm, premaxilla; pt, pterygoid; q, quadrate; so, supraorbital; sq, squamosal; v, vomers. Scale bar equals 100 mm.
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 (
(A) AMNH 22555 skull in lateral view; (B) Interpretative drawing of the photo in (A). (C) “
Morphology of cranial crests has been invariably used as a crucial character in the diagnoses of every single putative species of
At least one species assigned to
Our regression analysis, however, challenges the use of height and anteroposterior extension of the premaxillary crest as robust characters in the diagnosis of anhanguerids at the species level. As demonstrated here, anhanguerid skulls show statistically significant positive allometric growth of the premaxillary crest (see also the work of
Following the recent discovery of crested pterosaur assemblages preserving a large number of individuals belonging to a single species (
Naturally, these are characters used in a cladistic sense, but others have also been proposed as diagnostic of the genus. While comparing
Here we suggest the following revised diagnosis for
When first described by
Although AMNH 22555 is indeed similar to the “
Differences between AMNH 22555 and “
In spite of these remarkable differences between AMNH 22555 and the
A reappraisal of the purportedly diagnostic features of the individual
When first described,
(A)
“
“
Despite its large body size, the holotype of
Specimens attributable to
A possible overestimation in the anhanguerid diversity of the Romualdo Formation was also already pointed out by
As the relation between morphological disparity and speciation is vague, the application of the prevailing definition of the biological species concept (grounded on reproductive isolation) to the fossil record is exceedingly challenging (e.g.,
As we demonstrated, most of the allegedly diagnostic characters traditionally used to distinguish proposed
A natural ecological question that follows the assumption that Romualdo Formation pterosaur taxa were sympatric and coeval, is how such a large number of taxa with supposedly overlapping ecological niches may have coexisted. However, competitive exclusion of species happens only when the resources are scarce to the point of limiting population growth. If we assume, as is likely, that
Although our allometric regressions are not
The yet incipient results derived from controlled excavations on the Romualdo Formation already demonstrate clear evidence for faunal turnover, through the substitution of a basal fish assemblage dominated by the gonorynchiform
The temporal resolution of Romualdo Formation fossils was never estimated and several events of mass mortality probably took place (
Even though more than a dozen relatively complete skulls referable to the Anhangueridae and closely related taxa are nowadays held in public collections, this is the first study to perform a comprehensive morphometric analysis of continuous morphological features seen in the skulls of members of this clade. As a result, characters related to both dorsoventral height and the anteroposterior extension of the premaxillary crest are found to be allometrically correlated to skull size, and therefore at least in part to ontogeny. The observation that anhanguerid premaxillary crest morphology is size-dependent also means that it is largely unfit to be used as a diagnostic character for delimiting species, as has been commonly proposed for this group in the past. A taxonomic review excluding these characters reveals that as few as three
TPS file containing MorphoJ-friendly raw data for the performed morphometric analyses.
American Museum of Natural History, New York, USA
Museu de História Natural de Sintra, Sintra, Portugal
Museu Nacional/Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil
Museu de Paleontologia, Santana do Cariri, Brazil
Natural History Museum, London, UK
National Science Museum, Tokyo, Japan
Universidade do Estado do Rio de Janeiro, Rio de Janeiro, Brazil
National Natuurhistorisch Museum/Naturalis, Leiden, The Netherlands
Sammlung Oberli, a private collection belonging to Mr. Urs Oberli, Sankt Gallen, Switzerland
Staatliches Museum für Naturkunde, Karlsruhe, Germany
Staatliche Naturwissenschaftliche Sammlungen Bayerns/Bayerische Staatssammlung für Paläontologie und Geologie, Munich, Germany
For granting access to AMNH 22555 and other pterosaur specimens, the authors are particularly indebted to Mark Norell and Carl Mehling (AMNH). Also, we would like to thank the following people for allowing the study of specimens under their care and for kind help during visits: Álamo Saraiva and João Kerensky (MPSC); Oliver Rauhut and Markus Moser (SNSB-BSPG); Alexander Kellner and Helder Silva (MN); Eberhard Frey (SMNK); Rainer Schoch (SMNS); Sandra Chapman and Lorna Steel (NHMUK); Dan Pemberton and Matt Riley (CAMSM); Stephen Hutt (IWCMS); David Gelsthorpe (MANCH); Urs Oberli (SAO); Jon de Vos (Naturalis); Mauro Bon (MSN); and Wang Xiaolin (IVPP). FLP thanks Cesar Schultz for advisement on pterosaurs and Jose R.I. Ribeiro for initial advising in morphometric methodology, TR thanks Alexander Kellner for advisement and support. We also acknowledge Pedro Godoy (University of Birmingham) for comments and suggestions on the methodology applied here, Jonathan Tennant for suggestions and review of the language, as well as Renan Bantim and an anonymous referee for valuable comments on an early version of this paper.
The authors declare there are no competing interests.
The following information was supplied regarding data availability:
The raw data is included as a