The origin of turtles: a paleontological perspective

Chondrocranial anatomy of Testudo hermanni (Testudinidae, Testudines) with a comparison to other turtles

Using histological cross-sections, the chondrocranium anatomy was reconstructed for two developmental stages of Hermann's tortoise (Testudo hermanni). The morphology differs from the chondrocrania of most other turtles by a process above the ectochoanal cartilage with Pelodiscus sinensis being the only other known species with such a structure. The anterior and posterior processes of the tectum synoticum are better developed than in most other turtles and an ascending process of the palatoquadrate is missing, which is otherwise only the case in pleurodiran turtles. The nasal region gets proportionally larger during development. We interpret the enlargement of the nasal capsules as an adaption to increase the surface area of the olfactory epithelium for better perception of volant odors. Elongation of the nasal capsules in trionychids, in contrast, is unlikely to be related to olfaction, while it is ambiguous in the case of Sternotherus odoratus. However, we have to conclude that research on chondrocranium anatomy is still at its beginning and more comprehensive detailed descriptions in relation to other parts of the anatomy are needed before providing broad-scale ecological and phylogenetic interpretations.

Cranial ecomorphology of turtles and neck retraction as a possible trigger of ecological diversification

Turtles have a highly modified body plan, including a rigid shell that constrains postcranial anatomy. Skull morphology and neck mobility may therefore be key to ecological specialization in turtles. However, the ecological signal of turtle skull morphologies has not been rigorously evaluated, leaving uncertainties about the roles of ecological adaptation and convergence. We evaluate turtle cranial ecomorphology using three-dimensional geometric morphometrics and phylogenetic comparative methods. Skull shape correlates with allometry, neck retraction capability, and different aquatic feeding ecologies. We find that ecological variables influence skull shape only, whereas a key functional variable (the capacity for neck retraction) influences both shape and size. Ecology and functional predictions from three-dimensional shape are validated by high success rates for extant species, outperforming previous two-dimensional approaches. We use this to infer ecological and functional traits of extinct species. Neck retraction evolved among crownward stem-turtles by the Late Jurassic, signaling functional decoupling of the skull and neck from the shell, possibly linked to a major episode of ecomorphological diversification. We also find strong evidence for convergent ecological adaptations among marine groups. This includes parallel loss of neck retraction, evidence for active hunting, possible grazing, and suction feeding in extinct marine groups. Our large-scale assessment of dietary and functional adaptation throughout turtle evolution reveals the timing and origin of their distinct ecomorphologies, and highlights the potential for ecology and function to have distinct effects on skull form.

Common evolutionary origin of acoustic communication in choanate vertebrates

Acoustic communication, broadly distributed along the vertebrate phylogeny, plays a fundamental role in parental care, mate attraction and various other behaviours. Despite its importance, comparatively less is known about the evolutionary roots of acoustic communication. Phylogenetic comparative analyses can provide insights into the deep time evolutionary origin of acoustic communication, but they are often plagued by missing data from key species. Here we present evidence for 53 species of four major clades (turtles, tuatara, caecilian and lungfish) in the form of vocal recordings and contextual behavioural information accompanying sound production. This and a broad literature-based dataset evidence acoustic abilities in several groups previously considered non-vocal. Critically, phylogenetic analyses encompassing 1800 species of choanate vertebrates reconstructs acoustic communication as a homologous trait, and suggests that it is at least as old as the last common ancestor of all choanate vertebrates, that lived approx. 407 million years before present.

A new model of forelimb ecomorphology for predicting the ancient habitats of fossil turtles

Various morphological proxies have been used to infer habitat preferences among fossil turtles and their early ancestors, but most are tightly linked to phylogeny, thereby minimizing their predictive power. One particularly widely used model incorporates linear measurements of the forelimb (humerus + ulna + manus), but in addition to the issue of phylogenetic correlation, it does not estimate the likelihood of habitat assignment. Here, we introduce a new model that uses intramanual measurements (digit III metacarpal + non-ungual phalanges + ungual) to statistically estimate habitat likelihood and that has greater predictive strength than prior estimators. Application of the model supports the hypothesis that stem-turtles were primarily terrestrial in nature and recovers the nanhsiungchelyid Basilemys (a fossil crown-group turtle) as having lived primarily on land, despite some prior claims to the contrary.

The ecomorphology of the shell of extant turtles and its applications for fossil turtles

Turtles are a successful clade of reptiles that originated in the Late Triassic. The group adapted during its evolution to different types of environments, ranging from dry land to ponds, rivers, and the open ocean, and survived all Mesozoic and Cenozoic extinction events. The body of turtles is characterized by a shell, which has been hypothesized to have several biological roles, like protection, thermal and pH regulation, but also to be adapted in its shape to the ecology of the animal. However, only few studies have investigated the relationships between shell shape and ecology in a global context or clarified if shape can be used to diagnose habitat preferences in fossil representatives. Here, we assembled a three-dimensional dataset of 69 extant turtles and three fossils, in particular, the Late Triassic Proganochelys quenstedtii and Proterochersis robusta and the Late Jurassic Plesiochelys bigleri to test explicitly for a relationship between shell shape and ecology. 3D models were obtained using surface scanning and photogrammetry. The general shape of the shells was captured using geometric morphometrics. The habitat ecology of extant turtles was classified using the webbing of their forelimbs as a proxy. Principal component analysis (PCA) highlights much overlap between habitat groups. Discriminant analyses suggests significant differences between extant terrestrial turtles, extant fully aquatic (i.e., marine and riverine) turtles, and an unspecialized assemblage that includes extant turtles from all habitats, mostly freshwater aquatic forms. The paleoecology of the three fossil species cannot be determined with confidence, as all three fall within the unspecialized category, even if Plesiochelys bigleri plots closer to fully aquatic turtles, while the two Triassic species group closer to extant terrestrial forms. Although the shape of the shell of turtles indeed contains an ecological signal, it is overall too weak to uncover using shell shape in paleoecological studies, at least with the methods we selected.

Distribution of GFAP in Squamata: Extended Immunonegative Areas, Astrocytes, High Diversity, and Their Bearing on Evolution

Squamata is one of the richest and most diverse extant groups. The present study investigates the glial fibrillary acidic protein (GFAP)-immunopositive elements of five lizard and three snake species; each represents a different family. The study continues our former studies on bird, turtle, and caiman brains. Although several studies have been published on lizards, they usually only investigated one species. Almost no data are available on snakes. The animals were transcardially perfused. Immunoperoxidase reactions were performed with a mouse monoclonal anti-GFAP (Novocastra). The original radial ependymoglia is enmeshed by secondary, non-radial processes almost beyond recognition in several brain areas like in other reptiles. Astrocytes occur but only as complementary elements like in caiman but unlike in turtles, where astrocytes are absent. In most species, extended areas are free of GFAP—a meaningful difference from other reptiles. The predominance of astrocytes and the presence of areas free of GFAP immunopositivity are characteristic of birds and mammals; therefore, they must be apomorphic features of Squamata, which appeared independently from the evolution of avian glia. However, these features show a high diversity; in some lizards, they are even absent. There was no principal difference between the glial structures of snakes and lizards. In conclusion, the glial structure of Squamata seems to be the most apomorphic one among reptiles. The high diversity suggests that its evolution is still intense. The comparison of identical brain areas with different GFAP contents in different species may promote understanding the role of GFAP in neuronal networks. Our findings are in accordance with the supposal based on our previous studies that the GFAP-free areas expand during evolution.

Feeding biomechanics suggests progressive correlation of skull architecture and neck evolution in turtles

The origin of turtles is one of the most long-lasting debates in evolutionary research. During their evolution, a series of modifications changed their relatively kinetic and anapsid skull into an elongated akinetic structure with a unique pulley system redirecting jaw adductor musculature. These modifications were thought to be strongly correlated to functional adaptations, especially to bite performance. We conducted a series of Finite Element Analyses (FEAs) of several species, including that of the oldest fully shelled, Triassic stem-turtle Proganochelys, to evaluate the role of force distribution and to test existing hypotheses on the evolution of turtle skull architecture. We found no support for a relation between the akinetic nature of the skull or the trochlear mechanisms with increased bite forces. Yet, the FEAs show that those modifications changed the skull architecture into an optimized structure, more resistant to higher loads while allowing material reduction on specific regions. We propose that the skull of modern turtles is the result of a complex process of progressive correlation between their heads and highly flexible necks, initiated by the origin of the shell.

The evolution of dermal shield vascularization in Testudinata and Pseudosuchia: phylogenetic constraints versus ecophysiological adaptations

Studies on living turtles have demonstrated that shells are involved in the resistance to hypoxia during apnea via bone acidosis buffering; a process which is complemented with cutaneous respiration, transpharyngeal and cloacal gas exchanges in the soft-shell turtles. Bone acidosis buffering during apnea has also been identified in crocodylian osteoderms, which are also known to employ heat transfer when basking. Although diverse, many of these functions rely on one common trait: the vascularization of the dermal shield. Here, we test whether the above ecophysiological functions played an adaptive role in the evolutionary transitions between land and aquatic environments in both Pseudosuchia and Testudinata. To do so, we measured the bone porosity as a proxy for vascular density in a set of dermal plates before performing phylogenetic comparative analyses. For both lineages, the dermal plate porosity obviously varies depending on the animal lifestyle, but these variations prove to be highly driven by phylogenetic relationships. We argue that the complexity of multi-functional roles of the post-cranial dermal skeleton in both Pseudosuchia and Testudinata probably is the reason for a lack of obvious physiological signal, and we discuss the role of the dermal shield vascularization in the evolution of these groups. This article is part of the theme issue 'Vertebrate palaeophysiology'.

Comparative analysis of the shape and size of the middle ear cavity of turtles reveals no correlation with habitat ecology

The middle ear of turtles differs from other reptiles in being separated into two distinct compartments. Several ideas have been proposed as to why the middle ear is compartmentalized in turtles, most suggesting a relationship with underwater hearing. Extant turtle species span fully marine to strictly terrestrial habitats, and ecomorphological hypotheses of turtle hearing predict that this should correlate with variation in the structure of the middle ear due to differences in the fluid properties of water and air. We investigate the shape and size of the air-filled middle ear cavity of 56 extant turtles using 3D data and phylogenetic comparative analysis to test for correlations between habitat preferences and the shape and size of the middle ear cavity. Only weak correlations are found between middle ear cavity size and ecology, with aquatic taxa having proportionally smaller cavity volumes. The middle ear cavity of turtles exhibits high shape diversity among species, but we found no relationship between this shape variation and ecology. Surprisingly, the estimated acoustic transformer ratio, a key functional parameter of impedance-matching ears in vertebrates, also shows no relation to habitat preferences (aquatic/terrestrial) in turtles. We suggest that middle ear cavity shape may be controlled by factors unrelated to hearing, such as the spatial demands of surrounding cranial structures. A review of the fossil record suggests that the modern turtle ear evolved during the Early to Middle Jurassic in stem turtles broadly adapted to freshwater and terrestrial settings. This, combined with our finding that evolutionary transitions between habitats caused only weak evolutionary changes in middle ear structure, suggests that tympanic hearing in turtles evolved as a compromise between subaerial and underwater hearing.

Cerebrovascular β-dystroglycan immunoreactivity in vertebrates: not detected in anurans and in the teleosts Ostariophysi and Euteleostei

The aim of the present paper was to check for the presence of cerebrovascular dystroglycan in vertebrates, because dystroglycan, which is localized in the vascular astroglial end-feet, has a pivotal function in glio-vascular connections. In mammalian brains, the immunoreactivity of β-dystroglycan subunit delineates the vessels. The results of the present study demonstrate similar patterns in other vertebrates, except for anurans and the teleost groups Ostariophysi and Euteleostei. In this study, we investigated 1 or 2 representative species of the main groups of Chondrichthyes, teleost and non-teleost ray-finned fishes, urodeles, anurans, and reptiles. We also investigated 5 mammalian and 3 bird species. Animals were obtained from breeders or fishermen. The presence of β-dystroglycan was investigated immunohistochemically in free-floating sections. Pre-embedding electron microscopical immunohistochemistry on Heterodontus japonicus shark brains demonstrated that in Elasmobranchii, β-dystroglycan is also localized in the perivascular glial end-feet despite the different construction of their blood-brain barrier. The results indicated that the cerebrovascular β-dystroglycan immunoreactivity disappeared separately in anurans, and in teleosts, in the latter group before its division to Ostariophysi and Euteleostei. Immunohistochemistry in muscles and western blots from brain homogenates, however, detected the presence of β-dystroglycan, even in anurans and all teleosts. A possible explanation is that in the glial end-feet, β-dystroglycan is masked in these animals, or disappeared during adaptation to the freshwater habitat.

Slow and steady: the evolution of cranial disparity in fossil and recent turtles

Turtles (Testudinata) are a diverse group of amniotes that have a rich fossil record that extends back to the Late Triassic, but little is known about global patterns of disparity through time. We here investigate the cranial disparity of 172 representatives of the turtle lineage and their ancestors grouped into 20 time bins ranging from the Late Triassic until the Recent using two-dimensional geometric morphometrics. Three evolutionary phases are apparent in all three anatomical views investigated. In the first phase, disparity increases gradually from the Late Triassic to the Palaeogene with only a minor perturbation at the K/T extinct event. Although global warming may have influenced this increase, we find the Mesozoic fragmentation of Pangaea to be a more plausible factor. Following its maximum, disparity decreases strongly towards the Miocene, only to recover partially towards the Recent. The marked collapse in disparity is likely a result of habitat destruction caused by global drying, combined with the homogenization of global turtle faunas that resulted from increased transcontinental dispersal in the Tertiary. The disparity minimum in the Miocene is likely an artefact of poor sampling.

A Jurassic stem pleurodire sheds light on the functional origin of neck retraction in turtles

Modern turtles are composed of two monophyletic groups, notably diagnosed by divergent neck retraction mechanisms. Pleurodires (side-necked turtles) bend their neck sideways and protect their head under the anterior margin of the carapace. Cryptodires (hidden-necked turtles) withdraw their neck and head in the vertical plane between the shoulder girdles. These two mechanisms of neck retraction appeared independently in the two lineages and are usually assumed to have evolved for protective reasons. Here we describe the neck of Platychelys oberndorferi, a Late Jurassic early stem pleurodire, and find remarkable convergent morphological and functional similarities with modern cryptodires. Partial vertical neck retraction in this taxon is interpreted to have enabled fast forward projection of the head during underwater prey capture and offers a likely explanation to the functional origin of neck retraction in modern cryptodires. Complete head withdrawal for protection may therefore have resulted from an exaptation in that group.

Feeding behaviour in a 'basal' tortoise provides insights on the transitional feeding mode at the dawn of modern land turtle evolution

Almost all extant testudinids are highly associated with terrestrial habitats and the few tortoises with high affinity to aquatic environments are found within the genus Manouria. Manouria belongs to a clade which forms a sister taxon to all remaining tortoises and is suitable as a model for studying evolutionary transitions within modern turtles. We analysed the feeding behaviour of Manouria emys and due to its phylogenetic position, we hypothesise that the species might have retained some ancestral features associated with an aquatic lifestyle. We tested whether M. emys is able to feed both in aquatic and terrestrial environments. In fact, M. emys repetitively tried to reach submerged food items in water, but always failed to grasp them—no suction feeding mechanism was applied. When feeding on land, M. emys showed another peculiar behaviour; it grasped food items by its jaws—a behaviour typical for aquatic or semiaquatic turtles—and not by the tongue as generally accepted as the typical feeding mode in all tortoises studied so far. In M. emys, the hyolingual complex remained retracted during all food uptake sequences, but the food transport was entirely lingual based. The kinematical profiles significantly differed from those described for other tortoises and from those proposed from the general models on the function of the feeding systems in lower tetrapods. We conclude that the feeding behaviour of M. emys might reflect a remnant of the primordial condition expected in the aquatic ancestor of the tortoises.