Figure 4 - uploaded by Sérgio Ferreira Cardoso
Content may be subject to copyright.
Allometric trajectories among seven pangolin species (N = 241). A, the x-axis values are the log-transformed centroid sizes for each specimen; the y-axis values are the principal component 1 of the predicted values of a multivariate regression of shape on size. B-H, deformed meshes for the maximum (top) and minimum (bottom) shapes predicted from a multivariate Procrustes regression for each species are presented.
Source publication
Pangolins are among the most endangered groups of mammals, comprising eight extant species delineated into three genera. Despite several studies dedicated to their skeletal anatomy, the potential taxonomic insight from cranial morphological variation in extant Pholidota is yet to be assessed with modern geometric morphometric methods. We present th...
Contexts in source publication
Context 1
... intercepts of the regressions of predicted values of shape against size (Fig. 4) were relatively distinct within Phataginus and Manis, whereas both Smutsia species presented overlapping trajectories, as revealed by the Tukey comparisons (Supporting Information, Table S9). Considering the predicted shapes for minimum and maximum size resulting from the multivariate regression, the main size-related intraspecific ...
Context 2
... shapes for minimum and maximum size resulting from the multivariate regression, the main size-related intraspecific morphological change was the increase in length of the rostrum ( Figs 4, 5). Landmarks in the anterior part of the nasal and maxilla tended to be more anterodorsally positioned, and the nasal projected more posteriorly (e.g. Fig. 4, 5). The braincase was relatively lower in the adult, across all species, with the dorsal landmarks on the midline of the skull being more ventral when compared with their position in the juveniles (Fig. 5). Additionally, the landmarks placed on the zygomatic process of the maxilla and associated structures ( Fig. 3; landmarks 10, 57 ...
Context 3
... the pairwise matrix effect sizes were relatively small, and corresponding P-values were not significant. Therefore, all cryptic lineages presented parallel allometric trajectories ( Supporting Information, Fig. S4; Table S15). A Procrustes ANOVA revealed that cryptic lineages presented different skull shapes (F 4,65 = 3,12, P < 0.001; Table 2). ...
Context 4
... (13.0%) discriminated the WCA specimens (most positive values) from the remaining lineages (Fig. 7D). Group a posteriori probabilities retrieved a 75.4% attribution accuracy (see Supporting Information, Appendix S4). The vast majority of incorrect attributions were found in the major divisions WAF-GHA and CAF-WCA. ...
Context 5
... on this result, we performed an additional LDA with a priori attributions of WAF-GHA specimens to a western group (WES) and CAF-WCA specimens to an eastern group (CEN), while keeping DHG as a separate group (Fig. 8). Group a posteriori probabilities of the LDA of the three groups shows an attribution accuracy of 95.7% (see Supporting Information, Appendix S4). ...
Context 6
... explained a significant part of the total variation in skull shape among species within the Pholidota (evolutionary allometry; Table 1; Fig. 4; Supporting Information, Fig. S2; Table S2). The HOS and PTA tests showed that the directions of ontogenetic trajectories were conserved among pangolins, with only M. javanica and M. pentadactyla differing significantly from S. gigantea (HOS) and Tables S4 and S5). Ontogenetic trajectory angles did not differ within clades, which is ...
Context 7
... African and Asian clades clearly presented non-overlapping ontogenetic allometric trajectories for traits correlated with PC1 of the predicted values of multivariate regression of shape on size (Fig. 4), with African species sharing a higher intercept relative to Asian ones. We interpret this size-shape space as a good representation of the evolutionary allometry. Considering that parameters of modelled growth trajectories can be used efficiently as continuous characters in phylogenies ( Bardin et al., 2017), these different intercept ...
Context 8
... changes are better illustrated by the ontogenetic allometric trajectories of the two Smutsia species (Fig. 4; Supporting Information , Tables S6- S8), as they overlap in both size-shape and shape spaces ( Supporting Information, Fig. S1; Mitteroecker et al., 2005;Esquerré et al., 2017). Both PC1 of the predicted values and multivariate regressions suggest that S. temminckii is paedomorphic, because it tends to resemble juvenile S. gigantea ...
Context 9
... affinities. Differing ontogenetic allometric trajectories between closely related species were reported previously in hominids (Mitteroecker et al., 2004). Despite being sister taxa, humans and chimpanzees differ in skull shape from early ontogenetic stages. The allometric trajectories of the two Phataginus species exhibit a similar pattern ( Fig. 4; Supporting Information, Table S6). Both species show rather distinct intercepts (distinct shapes from early ...
Context 10
... our results suggest that complex allometric changes played an important role in the morphological evolution of the pangolin skull. All pangolins follow a similar ontogenetic trend characterized by the elongation of the rostrum and a posterior projection of the zygomatic process of the maxilla (Figs 4, 5). As a consequence, the braincase is relatively smaller in larger species. ...
Context 11
... small-sized Phataginus species are the only ones presenting shallow nasal notches ( Supporting Information, Fig. S9). The relative length of the parietal-squamosal suture (character 385; Gaudin et al., 2009) also appears to be influenced by allometry (Figs 4, 5). The parietalsquamosal suture is relatively longer in skulls with shorter snouts, which is the case in Phataginus species (the smallest pangolins) that show a relatively long parietal-squamosal suture. ...
Context 12
... parietalsquamosal suture is relatively longer in skulls with shorter snouts, which is the case in Phataginus species (the smallest pangolins) that show a relatively long parietal-squamosal suture. This is to be expected, because skull length is influenced mostly by elongation of the rostrum in larger species (Figs 4, 5). According to Gaudin et al. (2009), S. temminckii is the only species presenting a multistate for this character [< 25% greatest skull length (0) or > 25% (1)], which is congruent with its short snout and intermediate average size between that of Phataginus spp. ...
Context 13
... of the rostrum in larger species (Figs 4, 5). According to Gaudin et al. (2009), S. temminckii is the only species presenting a multistate for this character [< 25% greatest skull length (0) or > 25% (1)], which is congruent with its short snout and intermediate average size between that of Phataginus spp. and that of Manis spp. and S. gigantea (Fig. 4). These results therefore call for a revision of some characters included in morphological matrices used to reconstruct extant and extinct pangolin phylogenetic relationships. Based on morphological features, pangolins have been classified from a single genus (Manis) to a maximum of six different genera. In 1882, Jentink published a ...
Context 14
... for Phataginus; Gaubert et al., 2018), their skulls appear comparatively more distinct. Some of the differences between S. temminckii and S. gigantea are related to size (see above ; Fig. 4). The most extensive morphological phylogenetic work performed to date found that Smutsia was the least-supported modern genus, with only three unique unambiguous synapomorphies, none of which involve the cranium ( Gaudin et al., 2009). ...
Similar publications
We describe two partial skulls of juvenile individuals of Moradisaurus grandis, a moradisaurine captorhinid from the upper Permian Moradi Formation of northern Niger. The juvenile skulls are less than half the length of the largest known skull of M. grandis, and differ in featuring a transversely narrow mandible, only five rows of maxillary teeth,...
During the reproductive season, animals have to manage both their energetic budget and gamete stock. In particular, for semelparous capital breeders with determinate fecundity and no parental care other than gametic investment, the depletion of energetic stock must match the depletion of gametic stock, so that individuals get exhausted just after t...
Adelina papanoa sp. nov. from Mexico is described, illustrated and compared with the similar species Adelina bifurcata (Champion, 1893). The new species differs by the horns and depressions on the head and by the male genitalia.
Citations
... Resulting phylogenetic relationships would be based largely on functional morphology rather than deeply nested homologs. Convergence is common in nature (Carrano et al., 1999;Slater et al., 2009;Ferreira-Cardoso et al., 2019). Therefore, the work of comparative anatomists and taxonomists is to examine morphological trends and make inferences on which morphologies represent shared derived features representative of evolutionary relationships (homology) and convergently acquired features (homoplasy). ...
... 5,19). The congruent structures of the maxillary fenestra being maintained under various maxillary shapes suggests some level of modularity in the development of the maxillary fenestra (Fabbri et al., 2017;Ferreira-Cardoso et al., 2019). The morphological support for Acheroraptor having a close relationship with the deep snouted taxon Atrociraptor is reinforced by short branch lengths relative to other sister terminal taxa (Fig. 19). ...
Eudromaeosauria is a clade of derived dromaeosaurids that typifies the common perception of ‘raptor’ dinosaurs. The evolutionary history of this clade has been controversial due to conflicting views of taxonomic identity, and because, due to taphonomic bias, several species were diagnosed primarily or solely by the maxilla. The maxilla is therefore crucial in understanding the phylogenetic relationships within the clade. Morphometric characterization has been commonly applied to recognize and distinguish major dromaeosaurid clades. However, morphometrics mainly showed morphological convergence rather than phylogenetic relationships. This approach has made it difficult to get resolution of phylogenetic relationships among eudromaeosaurian taxa, often resulting in large polytomies or inconsistent placement of key species. To test previous character statements, computed tomography was used to analyze the maxillae of Acheroraptor, Atrociraptor, and Deinonychus, and compare them with other eudromaeosaurians from Asia and North America. Morphometric characters were examined, and regressions were used to look for allometric trends in maxillary dimensions and the relationship to topological landmarks within Eudromaeosauria and its outgroups. Characters were improved and implemented to better capture eudromaeosaurian morphological variation and better resolve their phylogenetic relationships. Phylogenetic analysis recovered three well-defined clades within Eudromaeosauria and corroborated occurrence data within the fossil record. Acheroraptor and Atrociraptor were recovered as derived members of Saurornitholestinae. Deinonychus is recovered as a basal eudromaeosaurian, sharing features with dromaeosaurines and saurornitholestines. These results challenge previous biogeographic hypotheses suggesting Asian and North American faunal interchange during the Late Cretaceous and support convergence of traits relating to snout dimensions and proportions.
... The taxonomy of pangolins and their relationship to other placental mammals have been debated extensively, though a range of molecular and combined morphological and molecular phylogenetic studies over the past several decades (e.g., Shoshani et al., 1985;Murphy et al., 2001a, b;O'Leary et al., 2013;du Toit et al., 2014;Hassanin et al., 2015;Gaubert et al., 2016Gaubert et al., , 2018Gaubert et al., , 2020 have helped establish Pholidota as the sister taxon to Carnivora and attribute extant pangolins to a single monophyletic family, Manidae. Though no morphological characters currently unite Pholidota and Carnivora , morphological and molecular phylogenetic analyses within Pholidota reach largely congruent results (e.g., Gaudin et al., 2009;Ferreira-Cardosa et al., 2020). ...
The fossil record for pangolins is sparse. Current biogeographic data suggest this clade originated in Europe, though pangolins seem to have disappeared from the European paleontological record during the middle Miocene, when they were hypothesized to have been pushed toward more tropical and sub-tropical equatorial environments due to global cooling trends. Here we report on a nearly complete humerus of a pangolin from the early Pleistocene (∼2.2–1.9 Ma) site of Grăunceanu, Romania. This fossil revises this previous understanding of pangolin evolution and biogeography and represents both the youngest fossil pangolin from Europe and the only fossil from Pleistocene Europe. The new species described here, Smutsia olteniensis, sp. nov., shares several synapomorphic traits with other Smutsia species, which are currently found only in Africa. However, relative to extant Smutsia, it has several unique traits that set it apart, including a longer and narrower entepicondyle, an enlarged supinator crest, and an enlarged greater tubercle. Together this unique suite of features justifies the description of a new species. This specimen definitively demonstrates that pangolins were present in Europe during the Pleistocene. Further, Smutsia has previously been thought to be an African taxon, with the oldest specimen from South Africa at ∼5 Ma and living species found across Africa. This specimen now demonstrates that Smutsia previously had a far larger biogeographic range. Finally, Grăunceanu has been reconstructed to have consisted of relatively open grasslands and woodlands, which is an unusual habitat for most pangolins.
A large part of extant and past mammalian morphological diversity is related to variation in size through allometric effects. Previous studies suggested that craniofacial allometry is the dominant pattern underlying mammalian skull shape variation, but cranial allometries were rarely characterized within cranial units such as individual bones. Here, we used 3D geometric morphometric methods to study allometric patterns of the whole skull (global) and of cranial units (local) in a postnatal developmental series of nine-banded armadillos (Dasypus novemcinctus ssp.). Analyses were conducted at the ontogenetic and static levels, and for successive developmental stages. Our results support craniofacial allometry as the global pattern along with more local allometric trends, such as the relative posterior elongation of the infraorbital canal, the tooth row reduction on the maxillary, and the marked development of nuchal crests on the supraoccipital with increasing skull size. Our study also reports allometric proportions of shape variation varying substantially among cranial units and across ontogenetic stages. The multi-scale approach advocated here allowed unveiling previously unnoticed allometric variations, indicating an untapped complexity of cranial allometric patterns to further explain mammalian morphological evolution.
Reliable identification of species is a key step to assess biodiversity. In fossil and archaeological contexts, genetic identifications remain often difficult or even impossible and morphological criteria are the only window on past biodiversity. Methods of numerical taxonomy based on geometric morphometric provide reliable identifications at the specific and even intraspecific levels, but they remain relatively time consuming and require expertise on the group under study. Here, we explore an alternative based on computer vision and machine learning. The identification of three rodent species based on pictures of their molar tooth row constituted the case study. We focused on the first upper molar in order to transfer the model elaborated on modern, genetically identified specimens to isolated fossil teeth. A pipeline based on deep neural network automatically cropped the first molar from the pictures, and returned a prediction regarding species identification. The deep-learning approach performed equally good as geometric morphometrics and, provided an extensive reference dataset including fossil teeth, it was able to successfully identify teeth from an archaeological deposit that was not included in the training dataset. This is a proof-of-concept that such methods could allow fast and reliable identification of extensive amounts of fossil remains, often left unstudied in archaeological deposits for lack of time and expertise. Deep-learning methods may thus allow new insights on the biodiversity dynamics across the last 10.000 years, including the role of humans in extinction or recent evolution.
The subject of this thesis is the morphological convergence in the skull of ant- and termite-eating placentals. Its goals are to characterize tooth reduction, covariance patterns, and morphological variation of the skull, and explore their link to the selective pressures associated to myrmecophagy.The first chapter focuses on the evolutionary, ontogenetic, and static variations of the skull in pangolins, a group of myrmecophagous animals that include the most threatened mammalian species on Earth. The morphological delimitation between seven of the eight species is demonstrated. Their ontogenetic allometric trajectories are described and the implications of the size variation on systematics are discussed. Additionally, intraspecific variation was partly associated to molecular distinctiveness of recently diverged cryptic species within the white-bellied pangolin. These results were obtained with the use of three-dimensional geometric morphometric methods.The second chapter was dedicated to the comparative anatomy of the mandible and masticatory apparatus. First, I investigate the internal mandibular anatomy on a comparative sample of placental mammals using µ-CT tomography and histology. Structures putatively associated to tooth innervation (dorsal canaliculi) are present in toothless anteaters, while they are absent in pangolins, which are equally toothless. Comparative anatomy performed intra- and interspecifically allowed to: i) show that dorsal canaliculi are invariably present in anteaters; ii) confirm the relationship between dorsal canaliculi and early tooth development; iii) show the independent evolution of dorsal canaliculi in anteaters and toothless whales. Dorsal canaliculi are vascularized and innervated in the collared anteater, despite its tooth loss. This suggests that despite tooth loss, tooth pulp innervation likely maintained its sensorial role on the dorsal part of the mandible of anteaters. The second part of chapter 2 is devoted to the comparative anatomy of the head musculature of the three extant anteater genera. Classical and digital dissections confirmed the reduction of the masticatory apparatus in anteaters. The masticatory apparatus of the pygmy anteater is found to significantly differ from that the other two genera. A comparison with the head musculatures of pangolins and aardvarks was done, based on previously published studies. Despite being myrmecophagous, the head musculature of aardvarks and pangolins shows some key differences from that of anteaters. This suggests that the feeding apparatus of ant- and termite-eating placentals varies at the functional level.The last chapter of this thesis covers the patterns of phenotypic covariance of the skull of 15 myrmecophagous species. A geometric morphometrics approach is used in order to explore and confirm hypotheses of modularity. Results show that patterns of modularity in myrmecophagous mammals resemble those of other placentals mammals. No common shift in the parcellation was found, other than that expected from the null hypothesis. Results suggest instead that skull elongation might have resulted on a slight remodeling of modularity patterns on the rostrum region in myrmecophagid anteaters. A preliminary analysis of ontogenetic trajectories of phenotypic covariance matrices in two myrmecophagous species shows that covariance patterns significantly change during ontogeny. This indicates that functional interpretations of static modularity and integration must be taken with caution.
The family Manidae includes all the living pangolins, or scaly anteaters. Pangolins are a small group, encompassing eight extant species that are distributed across sub-Saharan Africa, the Indian subcontinent, southern China, Southeast Asia, and the East Indies eastward to the Philippines. Four of these species presently occur in Africa: Smutsia gigantea, the giant pangolin; S. temmincki, the ground pangolin; Phataginus tetradactyla, the long-tailed pangolin; and P. tricuspis, the tree pangolin. All are found in the forests of western and central Africa except S. temmincki, which occurs in drier and more open habitats in eastern and southern Africa. The family Manidae is placed in the order Pholidota. Pholidota also includes at least two extinct families of pangolins, the Patriomanidae and Eomanidae. Additionally, several authors have advocated close ties between pangolins and an extinct group of mostly North American fossorial mammals with reduced dentitions, the palaeanodonts. This chapter describes the systematic paleontology of Pholidota.
