THE ANATOMICAL RECORD 295:1134–1146 (2012)
Skull Mechanics and the Evolutionary Patterns of the Otic Notch Closure in Capitosaurs (Amphibia: Temnospondyli) ´ -NOGUE ´ ,2 LLUIS GIL,2 JOSEP FORTUNY,1* JORDI MARCE 1 ` AND ANGEL GALOBART 1 Institut Catal a de Paleontologia Miquel Crusafont, Universitat Autonoma de Barcelona, Cerdanyola del Valle`s, Spain 2 Department of Strength of Materials and Structural Engineering, Universitat Polite`cnica de Catalunya, Terrassa, Spain
ABSTRACT Capitosaurs were among the largest amphibians that have ever lived. Their members displayed an amphibious lifestyle. We provide new information on functional morphology data, using finite element analysis (FEA) which has palaeoecological implications for the group. Our analyses included 17 taxa using (2D) plate models to test four loading cases (bilateral, unilateral and lateral bitings and skull raising system simulation). Our results demonstrates that, when feeding, capitosaurs concentrated the stress at the circumorbital region of the capitosaur skull and cranial sutures probably played a key role in dissipating and absorbing the stress generated during biting. Basal members (as Wetlugasaurus) were probably less specialized forms, while during Middle- and Late Triassic the group radiated into different ecomorphotypes with closed otic notch forms (as Cyclotosaurus) resulting in the strongest skulls during biting. Previous interpretations discussed a trend from an open to closed otic notch associated with lateral repositioning of the tabular horns, but the analysis of the skull-raising system reveals that taxa exhibiting posteriorly directed tabular horns display similar results during skull raising to those of closed otic notch taxa. Our results suggest that various constraints besides otic notch morphology, such as the elongation of the tabular horns, snout length, skull width and position, and size of the orbits affect the function of the skull. On the light of our results, capitosaur skull showed a trend to reduce the stresses and deformation during biting. Capitosaurs could be considered crocodilian analogues as they were top-level predators in fluvial and brackish Triassic ecosystems. Anat Rec, C 2012 Wiley Periodicals, Inc. 295:1134–1146, 2012. V
Key words: capitosauroidea; mastodonsauroidea; Temnospondyli; functional morphology; biomechanics; finite element analysis; otic notch
Additional Supporting Information may be found in the online version of this article. Grant sponsor: Synthesys grant; Grant number: FR-TAF-435. *Correspondence to: Josep Fortuny, Institut Catal a de Paleontologia Miquel Crusafont, Edifici ICP, Universitat Autonoma de Barcelona, Campus de Bellaterra s/n, 08193 Cerdanyola del Valle`s, Barcelona, Spain. E-mail:
[email protected] C 2012 WILEY PERIODICALS, INC. V
Received 28 April 2011; Accepted 4 April 2012. DOI 10.1002/ar.22486 Published online 9 May 2012 in Wiley Online Library (wileyonlinelibrary.com).
SKULL MECHANICS IN CAPITOSAURS
INTRODUCTION Capitosaurs, also known as mastodonsaurids (Damiani, 2001a), were members of the most diverse group of early tetrapods, the temnospondyls. Temnospondyls lived from the Carboniferous through the Early Cretaceous (Milner, 1990). During the Triassic, a group of temnospondyls, the Stereospondyli, radiated (Yates and Warren, 2000), achieving a worldwide distribution living in shallow marine, fluvial, brackish swamp, and terrestrial environments (Milner, 1990; Schoch and Milner, 2000; Warren, 2000). They displayed a significant amount of homoplasy, especially in the postcranium morphology (Defauw, 1989; Pawley, 2006). Capitosaurs were Triassic stereospondyls with a cosmopolitan range throughout Pangaea (Schoch and Milner, 2000). Previous geometric morphometrics work revealed a smaller disparity within the capitosaur clade in comparison with other stereospondyls, such as trematosauroids and rhydosteans (Stayton and Ruta, 2006). Capitosaurs were characterized by relatively large, heavy, parabolic skulls; an extensive pectoral girdle; and a tendency towards gigantism—reaching lengths of up to 6 m in some species, which were arguably among the largest amphibians that have ever lived (Schoch and Milner, 2000). They were secondarily aquatic animals exhibiting aquatic features such as a flattened skull, decreased bone ossification, and the presence of lateralline sensory canals (Fig. 1). Nevertheless, at least some capitosaur taxa probably retained a degree of terrestriality (see Sclerothorax in Schoch et al., 2007; Edingerella in Maganuco et al., 2009; and paracyclotosaurids in Mukherjee et al., 2010). Feeding and functional constraints in capitosaurs (and the whole group of Temnospondyli) have been mainly inferred from descriptions of their cranial morphology (e.g., Watson, 1958, 1962; Howie, 1970; Shishkin, 1987; Defauw, 1989; Schoch and Milner, 2000; Warren, 2000; Damiani, 2001a,b; Sulej and Majer, 2005; Jenkins et al., 2008; Damiani et al., 2009, and references therein). However, few studies have attempted to quantify of morphological trends related to cranial and postcranial shape variations within and across the early tetrapod clades and, particularly, in the Temnospondyli (e.g., capitosaurs) (i.e., Sengupta and Ghosh, 1993; Stayton and Ruta, 2006; Witzmann and Scholz, 2007; Witzmann et al., 2009; Sanchez et al., 2010). Various modes of carnivorous feeding have been postulated for temnospondyls (e.g., suction feeding or direct biting on prey) with some members, such as capitosaurs, have been hypothesized to be ‘‘crocodilomorph’’ analogues (Chernin and Cruickshank, 1978; Milner, 1990; Warren, 2000; Damiani, 2001b; Maganuco et al., 2009). Early hypotheses about the life mode of capitosaurs (Watson, 1958, 1962) considered them as mainly bottom-dwellers (i.e., benthic animals), which opened their mouths by the action of depressor mandibulae muscles. The contraction of these muscles could raise the skull because as bottom dwellers their jaws rested on the substrate. Howie (1970) contradicted this hypothesis; he considered the cleidomastoideus as the principal skull-raising muscle, with each muscle running between the dorsal process of the clavicle and a pronounced flange on the tabular, thereby offering a possible explanation for the lateral repositioning of the
1135
tabular horn during the evolutionary history of the group, which eventually resulted in the closure of the otic notch, while the mandible was lowered by the action of the depressor mandibulae. Subsequent works followed Howie’s approach and expanded upon his hypothesis, suggesting a more active lifestyle for capitosaurs. Supporting Howie’s conclusion, Chernin and Cruickshank (1978) argued that these animals showed anatomical adaptations which pointed to an active swimming lifestyle, characterizing them as mid-water predators analogous to modern crocodilians and rejecting suction feeding in capitosaurs. Suction feeding has been rejected for generalist aquatic tetrapod feeders (and might be wrong for capitosaurs as well) because of the impossibility of producing suction currents strong enough to surpass the escape speed of most prey (Taylor, 1987; Damiani, 2001b). More importantly, capitosaurs possessed akinetic skulls, unable to produce sufficient suction forces underwater (Schoch and Milner, 2000). As non-suction feeders, different modes of feeding could be considered such as jaw or lingual prehension, as in some extant amphibians (Deban and Wake, 2000), as well as direct biting, as in extant crocodiles (Busbey, 1995). Lastly, Damiani (2001b) proposed that capitosaurs were generalist aquatic top-level predators. He proposed that capitosaurs hunted prey by direct biting, using a rapid sideways sweep of the head during active swimming. This technique permitted the prey to be kept in sight as long as possible before striking. Here, we test the influence of cranial morphological differences among capitosaurs on skull mechanics by means of finite element methods (FEM). The primary objective of the present study is to analyze the interplay between shape, function, and evolution of structural complexes in the capitosaur clade such as the otic notch and the skull raising system. Plate finite element models were used with shell elements to evaluate 17 capitosaur taxa.
Phylogeny of the Clade Traditionally, the phylogeny of the capitosaurs was mainly based on changes in the squamosal embayment morphology, in particular its tendency to become a closed otic notch (Welles and Crosgriff, 1965; Ochev, 1966). Recent phylogenetic studies pointed out many more phylogenetically informative characters of the skull roof, palate, occiput, and mandibles (Schoch, 2000a, 2008; Damiani 2001a; Maganuco et al., 2009; Fortuny et al., 2011a). The phylogeny of capitosaurs is controversial with some genera and clades in debate. On this aim, the Early Triassic Edingerella is considered as either a lydekkerinid (Damiani 2001a), trematosaur (Schoch, 2008), or capitosaur (Steyer, 2003; Maganuco et al., 2009; Fortuny et al., 2011a). Here, we follow this last interpretation. On the other hand, different scenarios of cyclotosaurid origins are debated (See Damiani, 2001a; Schoch, 2008; Fortuny et al., 2011a for a discussion). In this study, we follow a recent phylogeny based on cranial characters (Fortuny et al., 2011a, see Fig. 2). This phylogeny does not include Sclerothorax hypselonotus, as its phylogenetic position within the basal members of the group is uncertain (Schoch et al., 2007; Schoch, 2008).
1136
FORTUNY ET AL.
Fig. 1. Dorsal and lateral skeleton reconstruction of a generalized capitosaur. Note the principal bones used in the present analysis to calculate its mechanical capabilities. M: Maxilla, PM: Premaxilla, QJ: Quadratojugal, SQ: Squamosal, T: Tabular.
The phylogeny in the work by Fortuny et al. (2011a) proposed Wetlugasaurus, Odenwaldia, and Vladlenosaurus as the most basal members of the Capitosauria, while the rest of capitosaurs were included in two large clades. On the first clade, Edingerella and Watsonisuchus were the most basal members including also more derived capitosaurs as Xenotosuchus, Cherninia, paracyclotosaurids (Stanocephalosaurus and Paracyclotosaurus), stenosaurids (Procyclotosaurus), and heylerosaurids (Quasicyclotosaurus and Eocyclotosaurus). On the second large clade, Parotosuchus was the most basal form with derived forms as Eryosuchus, Mastodonsaurus, Cyclotosaurus, and Tatatrasuchus (see Fortuny et al., 2011a for further details).
MATERIALS AND METHODS Model Construction Plate models of 17 taxa were analyzed (Tables 1 and 2 and Supporting Information Table S1). These taxa represent most of the skull variation found in the capitosaur clade. Plate models are not literal representations of reality but provide simplifications for the taxa that are being studied. The generation of models requires assumptions and the limitations of the models must be recognized. Herein, the capitosaur skull is not completely flat. Thus, it is not entirely reflective of the morphology of the capitosaur skull, but it can be used as an approximation for initial biomechanical investigations, and it can be tested through the use of more detailed 3D models in future study (Rayfield, 2004; Pierce et al., 2008, 2009, Fletcher et al., 2010). Assumptions have been applied (see below for details of each one): (a) All models were scaled to the same size.
To make comparison between different taxa, the same model size is required. Nevertheless, there is important intra- and interspecific size variability in capitosaurs during adult stage, with adults of same taxa showing important size differences (Supporting Information Table S1; Damiani, 2008). (b) Bone is considered as an elastic material without sutures and without considering rheological properties (Pierce et al., 2008, 2009). The role of sutures is out of the scope of the present study and should be analyzed in future 3D models. (c) Thickness of the model herein is arbitrary. Species models were created using the methodology outlined by Fortuny et al. (2011b). For the construction of the models, published drawings of well-known reconstructed skulls in dorsal view formed the basis of the biomechanical study. Reconstructions have been checked against original specimens wherever possible (indicated by an asterisk in Supporting Information Table S1). In other cases, reconstructions were compared with published photographs of the relevant specimens (if available). The steps for model construction, in summary, were as follows: (1) Each reconstruction was digitized in ImageJ v.1.36b software (http://rsb.info.nih.gov/ij/), developed at the US National Institute of Health (NIH). (2) Outline X, Y coordinates were imported into RhinocerosV v.4.0 software to create a planar surface. (3) The surface was improved and smoothed using SolidWorksV 2007. All the models were scaled to the same size (see below) in order to facilitate the comparison between them. (4) Placement of the fixing, muscle attachment, biting, and skull-raising landmarks was determined with MakeFan v.6.0 beta software (http://www3.canisius.edu/sheets/ morphsoft.html), developed by Sheets (2003), and imported to the surface model. (5) Each skull was R
R
SKULL MECHANICS IN CAPITOSAURS
1137
Fig. 2. Changes in skull mechanics in Capitosauria for selected taxa during Triassic. The upper figures correspond to FEA stress results during skull-raising system simulation while the lower figures correspond to FEA stress results during bilateral bite. The phylogeny is modified from Fortuny et al. (2011a) based on C Institute of Paleobiology, Polish Academy of Sciences. cranial characters. V R finite element analysis (FEA) analyzed using ANSYSV Package v.12.1 for Windows XP (32-bit system), in order to obtain the stresses and deformations of the model.
Model Properties Bone properties for extinct vertebrate models have been discussed since the first FEA analyses (see Rayfield, 2007 for a review). Temnospondyls have been hypothesized as the possible ancestors of lissamphibians (Ruta et al., 2007 and references therein). Unfortunately, the material properties available from extant amphibians are restricted to the femur (such as in Cryptobranchus and Ambystoma) (Erickson et al., 2002). On the other hand, capitosaurs were more heavily ossi-
fied than extant amphibians, which show strongly reduced ossification throughout the skeleton. Moreover, the microstructure of skull roof bones of capitosaurs show some similarities with amniotes (Witzmann, 2010) indicating that extant amphibians are a poor starting point for an analysis of capitosaur skeletal biomechanics. Consequently, the cranial materials properties of the unrelated, but more closely analogous extant crocodiles were used. Elastic properties were assumed for the models, using the following values: E (Young’s modulus): 6.65 GPa and m (Poisson’s ratio) 0.35. These values for Crocodylus frontal and prefrontal bones are based on the work by Currey (1987). As Rayfield and Milner (2008) pointed out, further understanding of the patterns of material property variation in extant taxa is
9.4012 3.2742 8.1237 4.318 6.4956 4.6307 6.3313 7.1202 6.0255 7.867 7.3786 7.8347 6.5393 4.1259 5.6725 3.9847 3.6344
11.376 5.8842 11.51 6.2252 8.5081 6.1168 7.7475 8.3128 7.4893 8.6204 11.576 8.9193 7.0853 5.284 6.1895 4.8564 4.9032
Maximum Orbital Von Mises stress (MPA) (10 1) 16.961 7.6302 18.952 10.21 12.335 15.734 11.673 14.057 9.5875 13.574 8.4036 9.1439 10.031 2.594 5.32 4.2055 2.7982
Maximum Otical Von Mises stress (MPA) (10 1) 6.6872 2.3959 6.007 3.8776 4.4612 3.437 4.498 4.6587 4.3 4.3595 4.5318 5.3722 4.577 2.7125 3.8094 3.0404 2.6778
Maximum deformation (mm) (10 1) 7.923 4.2272 8.1673 4.8451 6.0746 4.224 5.5723 5.7611 5.1522 4.4258 4.984 5.2771 4.5821 3.7189 4.129 3.4941 3.3796
Maximum Orbital Von Mises stress (MPA) (10 1)
Unilateral bite
13.71 7.6966 19.15 8.7966 11.427 16.47 10.114 13.483 8.6365 12.211 9.6483 8.3992 8.3416 2.8913 5.6325 4.3505 2.9734
Maximum Otical Von Mises stress (MPA) (10 1)
Wetlugasaurus angustifrons Parotosuchus orenburgensis Edingerella madagascariensis Sclerothorax hypselonotus Watsonisuchus aliciae Odenwaldia heidelbergensis Eryosuchus garjainovi Cherninia denwai Xenotosuchus africanus Stanocephalosaurus pronus Paracyclotosaurus davidi Mastodonsaurus giganteus Tatrasuchus wildi Eocyclotosaurus woschmidti Procyclotosaurus stantonensis Quasicyclotosaurus campi Cyclotosaurus robustus
Species 0.19226 0.132 0.11487 0.05942 0.19978 0.1902 0.14774 0.20611 0.14416 0.19764 0.19716 0.20886 0.14181 0.30483 0.19268 0.15645 0.1052
Maximum deformation (mm) (10 3) 3.1291 1.9447 1.8449 1.5519 2.4147 2.4127 2.3781 1.3264 1.4503 2.2269 2.2482 4.8072 1.7474 4.0816 2.0591 2.6544 1.7891
Maximum Orbital Von Mises stress (MPA) (10 3)
Lateral case
3.4374 2.0992 2.7767 1.8066 2.5083 4.4706 3.2429 1.1316 1.8338 2.4303 3.1455 2.9921 2.9817 3.5099 3.7311 5.791 1.4532
Maximum Otical Von Mises stress (MPA) (10 3)
10.61 8.5721 11.302 19.095 12.168 12.033 21.55 17.03 25.448 15.705 14.08 16.423 22.947 8.298 14.9 14.419 9.8571
Maximum deformation (mm) (10 3)
1.7672 1.486 1.7637 2.2676 2.0043 1.7018 2.2709 3.6994 3.0577 3.3996 2.255 1.8226 2.5871 1.7044 1.5542 1.916 2.0345
Maximum Orbital Von Mises stress (MPA) (10 3)
Skull raising
4.3938 6.351 4.08 8.2946 5.5641 4.7847 8.0726 7.0613 9.1734 5.8926 6.0166 6.307 8.7914 3.4111 5.7442 4.7075 4.2373
Maximum Otical Von Mises stress (MPA) (10 3)
TABLE 2. Maximum deformation and von mises stress in lateral and skull raising load case for the capitosaur species studied in the present work
Wetlugasaurus angustifrons Parotosuchus orenburgensis Edingerella madagascariensis Sclerothorax hypselonotus Watsonisuchus aliciae Odenwaldia heidelbergensis Eryosuchus garjainovi Cherninia denwai Xenotosuchus africanus Stanocephalosaurus pronus Paracyclotosaurus davidi Mastodonsaurus giganteus Tatrasuchus wildi Eocyclotosaurus woschmidti Procyclotosaurus stantonensis Quasicyclotosaurus campi Cyclotosaurus robustus
Species
Maximum deformation (mm) (10 1)
Bilateral bite
TABLE 1. Maximum deformation and von mises stress in bilateral and unilateral load case for the capitosaur species studied in the present work
1138 FORTUNY ET AL.
SKULL MECHANICS IN CAPITOSAURS
1139
the maximum values were recorded for the Von Mises stress in the otic notch and orbital area. Each step of the convergence procedure represents an increment in the number of nodes and elements of the skull mesh. Figure 3 corresponds to the convergence norm partially applied only in the otical and orbital areas in the case of Cyclotosaurus in a bilateral bite case. It is shown that in each step, a refinement of the mesh appears in the regions close to the otical and orbital areas (most interesting areas for applying the convergence of the mesh).
Boundary and Loading Conditions
Fig. 3. Finite Element Analysis (FEA) methods. A: Loading conditions on the skull model in the four cases considered. B: Convergence procedure of the Cyclotosaurus mesh in the otical and orbital area.
needed in order to make justifiable assumptions as to direction, and orthotropy in cranial bone. Potential orthotropic properties of the capitosaur skull were, therefore, not included because the data are not yet available. A mesh of rectangular and triangular shell elements was created, and a uniform thickness of 2 mm was assumed for each model. The thickness is arbitrary and the application in all the models implies that the no consideration of its influence allows a comparison between the different models. A starting refined mesh is automatically generated by the FEA package, and convergence tools were used as part of the solution process (h-adaptive method). The FEA package controls the level of accuracy for selected results of stress and employs an adaptive solver engine to identify and refine the model in areas that benefit from adaptive refinement. The convergence norm for stress in structural analysis (Zienkiewicz and Zhu, 1992a,b) was used to control the results of the element and its convergence. The convergence procedure of the mesh and how the mesh changed in each step is shown in Fig. 3. The key factors that drive to the convergence are the mesh density and the element distribution rather than the number of nodes and elements themselves. On this aim,
FEA was carried out for four separate load cases (bilateral, unilateral, lateral, and skull raising) (Fig. 3) on all the models in order to quantify the stress response in relation to the loading conditions. We did not attempt to determine the realistic stress and deformation values in the models, but used relative values to compare skull morphologies within capitosaurs. To accomplish this goal we applied a unit force because the material is elastic (Dumont et al., 2009). The adductor jaw musculature of temnospondyls was mainly attached to the cheekbones, particularly the squamosal (Carroll and Holmes, 1980; Schoch and Milner, 2000; Witzmann and Pfretzschner, 2003; Carroll, 2007). The squamosal was used as the adductor musculature origin for the bilateral and unilateral loading cases. The area of squamosal force was determined by landmarks placed in the junctions between the squamosal and neighboring tabular, supratemporal, postorbital, jugal, and quadratojugal. In these four cases, the models were simply supported (displacement in X, Y, Z constrained and free rotation along the axes) along the posterior edge of the left and right quadratojugal during loading. We recorded only the relative differences in stress between two areas: otic and orbital, when size is held constant. In our 2D approach, bending was the main mechanism of performance. Deformation by bending is a relative measure of deflection between points. Hence our concern was basically to keep the relative distance among points during the scale of the geometry for each skull. To do this, we used a homothetic projection (isometric transformation of coordinates) and we took a pattern length as a comparative among skulls. In the tables, we multiplied the results in order to facilitate their interpretation (Tables 1 and 2). To avoid artificial noise created by the simply supported boundary conditions, especially in the premaxilla-maxilla suture, the maximum Von Mises stress value in the skull was not recorded. The maximum value appears artificially at points where the boundary conditions retain displacements as a simple support. The mesh convergence tests done for each model validate that the stresses in this area follow a vertical asymptotic tendency to increase the value towards infinity. This numerical singularity is well-known in the field of FEA and this assumption can drive a partial analysis of results because mistakenly high stresses are generated at constrained nodes, and therefore results from these areas should not be considered. In this study, the results are focalized to record the maximum Von Mises stress value in the areas of interest: the otic and orbital areas, which are suitable to converge. Therefore, we omitted
1140
FORTUNY ET AL.
Fig. 4. Deformation and Von Mises magnitudes and distribution in Wetlugasaurus, Parotosuchus, Eocyclotosaurus, Mastodonsaurus, and Cyclotosaurus in bilateral (top) and unilateral (bottom) bite simulations. Units are mm in the deformation colour bar and MPa in the stress colour bar.
the singularity provided by the maximum Von Mises value at the boundary conditions region. Numerically, it is possible to overcome the artificial noise produced by the singularity in the simply supported boundary condition by using a very coarse mesh in this area. The low accuracy of this coarse mesh eliminates the artificial noise, while the Saint-Venant principle assures that the values of deformation can still be recorded. The Von Mises criterion is an isotropic criterion traditionally used to predict the yielding of ductile materials such as metals. Bone could be considered brittle (Doblare et al., 2004) or ductile (Dumont et al., 2009) but, according to Doblare et al. (2004), when isotropic
material properties are used in cortical bone, the Von Mises criterion may be the most accurate for predicting fracture location regardless its consideration. Color-coded stress distribution shows the pattern of peak stress in the capitosaur skull during bilateral, unilateral, and lateral loading of the snout, and also in skull-raising (Figs. 4 and 5 and Supporting Information Figs. S1–S4). The total deformation is always an appropriate value to record because it is without artificial noise and without considerations of criterion. Bilateral bite (Fig. 4): Simulates a bilateral bite at the left and right premaxilla-maxilla suture, with 1 N applied to each side in the perpendicular direction of the
SKULL MECHANICS IN CAPITOSAURS
1141
Fig. 5. Deformation and Von Mises magnitudes and distribution in Wetlugasaurus, Parotosuchus, Eocyclotosaurus, Mastodonsaurus, and Cyclotosaurus in lateral bite simulation (top) and skull-raising system (bottom). Units are mm in the deformation colour bar and MPa in the stress color bar.
plate, to bend the skull dorsally. The bite force was applied at the adductor jaw musculature, placed in the squamosal. A simply supported boundary condition was applied in the premaxilla-maxilla suture (approximately in the position of the fangs present in the palate). Unilateral bite (Fig. 4): Simulates a unilateral bite at the left premaxilla-maxilla suture with 1 N applied in the z-direction to induce bending and superimposed torsional loading. The bite force was applied in the adductor jaw musculature, placed in the squamosal. A simply supported boundary condition was applied in the premaxilla-maxilla suture (approximately in the position of the fangs present in the palate).
Lateral bite (Fig. 5): Simulates a lateral loading at the left premaxilla-maxilla suture with 1 N applied in the ydirection to generate a within-plane lateral bend to the snout. The force was applied in the position of the fangs present in the palate. The purpose of this case is to simulate movement of the head through the water in order to test Damiani’s (2001b) hypothesis, considering that capitosaurs could hunt prey by using a rapid sideways sweep of the head during active swimming. Skull-raising system case (Fig. 5): Simulates a bilateral skull-raising system at the tabular horns applied in the y-direction to induce the elevation of the skull. Considering the cortical bone density to be 2 g/cm3, the
1142
FORTUNY ET AL.
weight of the skull is not needed to be calculated. As cortical bone density is estimated, a pressure in all the surface of the skull with this pressure’s value is applied. A simply supported boundary condition was applied in the tabular horn, considering on one hand, the triple suture between the tabular, the post-parietal, and the supratemporal, and on the other hand, the posterior edge of the tabular. The three first load simulations are designed to reflect dorsoventral, torsional, and mediolateral loads. The fourth simulation is designed to evaluate the hypothesis that the closure of the otic notch through a lateral reorientation of the tabular horn, a key character of the group, was evolved to resist stresses produced during jaw depression (Howie, 1970).
RESULTS The standard bite for the extinct group of capitosaurs is unknown. Distribution and values of equivalent Von Mises stresses and total deformations (Tables 1 and 2 and Figs. 4 and 5, Supporting Information Figs. S1–S4) are recorded for each skull in order to compare their behavior under the effect of the loads and constraints as defined for the four cases. The numerical results of deformation and stress are presented in Tables 1 and 2.
Biting Simulations The skull deformation pattern during bilateral and unilateral biting (Fig. 4, Supporting Information Figs. S1 and S2) peaks in the orbital area and/or in the preorbital zone. Maximum deformation values in bilateral and unilateral simulations are found in Edingerella madagascariensis and Wetlugasaurus angustifrons (with values from 10 to 9 mm in bilateral and 8 to 6 mm in unilateral cases). Lowest deformation (less than 4 mm in bilateral and 3 mm in unilateral cases) are found in Parotosuchus orenburgensis and Cyclotosaurus robustus. Stress patterns during bilateral and unilateral bites show that stress increases through the skull in an anteroposterior direction during biting—peaking around the orbits and otic notch. In the bilateral case, an examination of stress magnitude in the orbital area shows Edingerella madagascariensis, Paracyclotosaurus davidi, and Wetlugasaurus angustifrons to be the weakest, with values ranging around 11 MPa. In contrast, the strongest skulls, as recorded by Von Mises stress in orbital area, ranges around 5 MPa in Quasicyclotosaurus and Cyclotosaurus. Results of the otic notch area in the bilateral case shows that Edingerella madagascariensis, Wetlugasaurus angustifrons, and Odenwaldia heidelbergensis (ranging from 19 to 15 MPa) have the weakest skulls. The strongest skulls are found in Cyclotosaurus robustus and Eocyclotosaurus lehmani, with values ranging from 5 to 2 MPa. The stress magnitude in the unilateral case at the orbital area reveals Edingerella madagascariensis, Wetlugasaurus angustifrons, and Watsonisuchus aliciae to have the weakest skulls (ranging from 8 to 6 MPa). The three genera with closed otic notches (Cyclotosaurus, Eocyclotosaurus, and Quasicyclotosaurus) have the strongest skulls in the orbital area, ranging less than 4 MPa. Von Mises stress recorded in the otic notch area in the unilateral cases shows Edingerella madagascariensis and Odenwaldia
heidelbergensis to have the weakest skulls (ranging from 17 to 16 MPa), while the strongest skulls are represented by the three above mentioned closed otic notch genera with values below the 3 MPa. On the other hand, during lateral loading of the snout (Fig. 5 and Supporting Information Fig. S3), the skull deformation pattern peaks in the anterior area surrounding the naris. Maximum deformation values are found in Eocyclotosaurus lehmani and Mastodonsaurus giganteus (with values from 0.3 to 0.2 mm). The lowest deformation values are found in Sclerothorax hypselonotus and Cyclotosaurus robustus with values below 0.11 mm. Stress patterns during lateral loading show an increase of stress along the posterior margin of the skull, especially increasing in the orbital region (including the jugal). The examination of stress magnitude in the orbital area shows Mastodonsaurus giganteus, Eocyclotosaurus lehmani, and Wetlugasaurus angustifrons to have the weakest skulls during lateral loading, with values ranging from 4.8 MPa to 3 MPa. The strongest skulls are represented by Cherninia denwai and Xenotosuchus africanus with values nearby 1.5 MPa. Finally, Von Mises stress recorded in the otic notch during lateral loading results in the finding that Quasicyclotosaurus campi and Odenwaldia heidelbergensis have the weakest skulls, with values above the 4 MPa. The strongest skulls, ranging below 1.4 MPa, are represented by Cherninia denwai and Cyclotosaurus robustus.
Skull-Raising Simulation The skull deformation pattern in the skull-raising simulation (Fig. 5 and Supporting Information Fig. S4) peaks in the snout, including the maxilla, nasal, and premaxilla bones. Maximum deformation values are found in Xenotosuchus africanus and Tatrasuchus wildi, with values ranging from 22 to 25 mm. In contrast, the lowest deformation values are found in Eocyclotosaurus lehmani and Parotosuchus orenburgensis, ranging around the 8 mm. Von Mises stress recorded in otic and orbital areas peaks at the anterior part of the tabular bone, including in some cases, the squamosal bone. The weakest skulls, as recorded in the orbital area are seen in Cherninia denwai and Stanocephalosaurus pronus with values above the 3.3 Mpa. Contrastingly, the strongest skulls are represented by Parotosuchus orenburgensis and Procyclotosaurus stantonensis with orbital values ranging nearby 1.5 MPa. Finally, the Von Mises record on the otic notch area reveals that Eocyclotosaurus lehmani and Edingerella madagascariensis have the weakest skulls with values below 4 MPa, while the skull of Xenotosuchus africanus and Tatrasuchus wildi are the strongest, with values above 8.7 MPa.
DISCUSSION Mechanical Performance During Biting Loads The terrestrial mode of feeding (direct bite on prey) first emerged in aquatic taxa, as demonstrated by the analysis of the cranial sutures of the early tetrapod Acanthostega (Markey and Marshall, 2007). Cranial sutures are important indicators of skull function (Rafferty and Herring, 1999). The dynamic stability of the cranial bones represents a compromise between
SKULL MECHANICS IN CAPITOSAURS
strength, compression, and torsional forces, and the requirement to economize energy to maintain cranial bone structures (Lanyon and Rubin, 1985; Kathe, 1999). Two main functional factors for sutures should be considered: energy absorption and stress flexion. To date, up to eight different sutural types have been described in temnospondyls and archosaurs (Kathe, 1995, 1999; Busbey, 1995). As demonstrated by our FEA results, the circumorbital region in capitosaurs underwent a large amount of stress, confirming previous interpretations (Bolt, 1974). This is in agreement with the great number of sutures in the circumorbital region which were able to absorb and flexion forces (Kathe, 1999). Nevertheless, future works with 3D models should analyze in more detail on the function of the sutures during the different biting simulations. A recent work (Fortuny et al., 2011b) showed that capitosaurs were amphibious feeders while other temnospondyls could be considered terrestrial feeders (as edopoids) or fully aquatic feeders (as the case of some archegosaurs and trematosaurs). The results herein presented enable us to deepen on the feeding ecology of these amphibious feeders and suggest a more diverse range of ecomorphotypes than previously recognized for capitosaurs. FEA revealed that the capitosaur clade underwent important mechanical changes over time, affecting the feeding capabilities of these amphibious tetrapods. Damiani (2001b) hypothesized that the capitosaurs were generalist aquatic top-level predators, with a putative diet of fish, amphibians, and small reptiles. However, he did not rule out the possibility that the presence of massive tusks, deep mandibles, and jaw musculature would have allowed for larger prey. FEA models (Figs. 4 and 5 and Supporting Information Figs. S1–S3) allow us to examine Damiani’s hypothesis more closely. Our results show that basal members of the clade (Fig. 2), such as the Early Triassic Edingerella madagascariensis and Wetlugasaurus angustifrons, possessed the weakest skulls according to the bi- and unilateral cases, and possibly represent less specialized forms. On this aim, basal stereospondylomorphs (e.g., Archegosaurus, Sclerocephalus, and Glanochthon) possessed also weak skulls during bi- and unilateral cases (see Fortuny et al., 2011b for further details). This is especially important for capitosaurs as its members shows an important decreasing of stress values from basal to more derived forms in contrast with other stereospondyls as trematosaurs and archegosaurs with high mean stress values (Fortuny et al., 2011b). Possibly, the otic notch morphology plays an important role in these changes on the stress values as the closed otic notch genera capitosaurs such as Cyclotosaurus, Eocyclotosaurus, and Quasicyclotosaurus represented the strongest skulls under the same simulated conditions. But this tendency from open to closed otic notch should be approached with caution; the cosmopolitan genus Parotosuchus displays posteriorly directed tabular horns, but lack an embayment. Nevertheless, it shows similar results to the closed otic notch and represents one of the strongest and most stable skull morphologies. These results suggest that the mechanical performance for Parotosuchus could explain the widespread distribution of this genus during the Early Triassic, comprising at least eight valid species, and reaching body lengths of about 3 m (Schoch and Milner, 2000; Fortuny et al.,
1143
2011a, Fig 1). This implies that otic notch morphology is not the only factor affecting the functional morphology of the capitosaur skull: elongation of the tabular horns (Sulej and Majer, 2005), snout length, skull width, and orbit position and size are also important. Considering other amphibious feeders, the Permian sclerocephalids also show open otic notch feeding in lake habitats (Schoch, 2009; Fortuny et al., 2011b) but with important amount of stress during bi- and unilateral bites in comparison with Parotosuchus or closed otic notch capitosaurs. In contrast, some forms of edopoids (as Nigerpeton) resulted in similar stress values when feeding, although they probably fed mostly on land (Fortuny et al., 2011b). During the Middle Triassic, the capitosaurs radiated (Fig. 2). Some forms recorded in Europe, such as Mastodonsaurus and Eryosuchus, attained gigantic size (Schoch, 1999, 2000b). Others, such as the stenosaurins and cyclotosaurid-related forms also flourished in Europe (Fortuny et al., 2011a). On the other hand, Xenotosuchus and Cherninia have been only described from Gondwana (South Africa, Zambia, and India) (Damiani, 2001a), while paracyclotosaurids have been discovered in more areas of Gondwana (Australia, India, and South Africa) and Laurasia (North America) (Damiani, 2001a). Heylerosaurins are known from North America and Europe, with a widespread distribution in the case of Eocyclotosaurus. Middle Triassic taxa displayed differences in their feeding ecology. The gigantic genus Mastodonsaurus proved to have one of the weakest skulls during the different biting simulations. Nevertheless, it should be noted that all the skulls were scaled at the same size in the present analysis. Several skulls has been recovered for Mastodonsaurus, reaching more than 65 cm length (Schoch, 1999), as similarly also happens in Eryosuchus (Schoch and Milner, 2000). Thus, its relatively large skull probably was capable of withstanding larger absolute forces than some other smaller capitosaurs, such as the closed otic notch forms or others such as the stenosaurine Procyclotosaurus. Other genera such as Paracyclotosaurus and Stanocephalosaurus displayed similar hardiness in bi- and unilateral simulations, corroborating the differences in feeding capabilities with cyclotosaurids, as previously suggested (Sulej and Majer, 2005). On the other hand, Eocyclotosaurus and Cherninia were considered as morphospace outliers as a result of their elongated snout, abbreviated skull table, and closely spaced orbits (Stayton and Ruta, 2006). Interestingly, the FEA lateral load case exhibited important stress differences between these two genera (see Table 2). Our mechanical results for Eocyclotosaurus suggest that its skull morphology is an adaptation to reduce drag during lateral movement of the head, increasing the efficiency to capture small rapid prey, as happens in extant aquatic predators with long and narrow snouts (Taylor, 1987), while Cherninia appears to be less effective for rapid sideways sweep of the head during active swimming. During the Late Triassic, only two capitosaur genera were present (Fig. 2), Cyclotosaurus and Mastodonsaurus, displaying important differences in deformation and stress distribution as discussed above. These mechanical differences are consequences of the more acute snout morphology in Mastodonsaurus in comparison with Cyclotosaurus, and differences in orbit size and otic
1144
FORTUNY ET AL.
notch morphology. Although they were generalist predators, our results indicate different mechanical capabilities, as the skull of Mastodonsaurus was more agile for rapid sideways sweep of the head, although the bites were weaker in comparison with Cyclotosaurus of the same size. Nevertheless, as previously mentioned, the gigantic sizes acquired for Mastodonsaurus might imply that the absolute forces were similar (or higher) than other capitosaurs with stronger results. Future research in capitosaur functional morphology using 3D models should refine bite analyses. In particular, examining the role of the hyobranchial skeleton as well as expand upon the importance of the cranial sutures (Kathe, 1999; Rayfield, 2005), and the mechanical consequences of the ontogenetical changes (Steyer, 2000; Damiani, 2008).
Skull-Raising System FEA results (Fig. 5 and Supporting Information Fig. S4) reveal a deformation pattern peaking along the snout while the stress pattern is close to the otic notch. Interestingly, these patterns were higher when tabular horns were posterolaterally or lateral directed, as was the case in Tatrasuchus and Xenotosuchus. In contrast, taxa displaying posteriorly directed tabular horns (Wetlugasaurus), and especially genera with a closed otic notch (Cyclotosaurus), showed less deformation. Maximum stress was broadly placed on the anterior part of the tabular, including the tabular-postparietal and tabular-squamosal sutures (and supratemporal, when this bone entered the dorsal margin of the otic notch). These data suggest that the possible explanation for the lateral reorientation of the tabular horn, which eventually closed the otic notch in some forms, as hypothesized by Howie (1970), should be approached with caution. An evolutionary tendency to enclose the otic notch appears to be correlated with posterolaterally or laterally directed tabular horns, arguably because the closed otic notch offered an optimal solution for deformation and stress distribution during skull raising. However, posteriorly directed tabular horns produced similar results to those of taxa with closed otic notches, falsifying this hypothesis as advanced by Howie (1970). Therefore, the cause (or causes) which drive the first changes in otic notch morphology remains an open question. Although capitosaurs were generalist predators, some degree of diet specialization was present in the group (see above). Perhaps, these differences in dietary preferences and feeding strategies were the key evolutionary processes that caused the differences in the skull-raising system and biting loads and, as a consequence, the changes in otic notch morphology (and other characters previously discussed). After the Permian-Triassic extinction, continental ecosystems were rebuilt during Early–Middle Triassic (Benton et al., 2004). Several ecological niches were left vacant, facilitating the diversification of the groups in the freshwater-brackish ecosystems. Capitosaurs possibly diversified, taking advantage of the unoccupied niches (as the niche previously occupied in lakes by the Permian sclerocephalids with also probably amphibious feeding (Fortuny et al., 2011b)) with different proportions of prey available (i.e., fishes versus small-to-medium-sized reptiles) as modern crocodiles do (Eusuchia: Crocodylia, excluding gavialids and tomisto-
mins) (Cleuren and De Vree, 2000). As an example of this diversification, cyclotosaurids presented a tendency to approach tabular and squamosal bones evolved to an embayment in contrast with the basal forms with open otic notch while other forms evolved to gigantic sizes with usually rounded snout and proportionally larger orbits, as the case of Mastodonsaurus and Eryosuchus enabling them to hunt bigger preys. Interestingly, the tendency to enclosure the otic notch occurred at least in two lineages of capitosaurs cyclotosaurids (e.g., Cyclotosaurus) and heylosaurins (Eocyclotosaurus) (Fig. 2). The presence of two lineages with convergent evolution with an embayment on the otic notch reinforces the optimal mechanical solution to reduce stress and deformation values during the skull raising. On one hand, the Anisian (Middle Triassic) heylosaurins (Eocyclotosaurus, Quasicyclotosaurus, and Procyclotosaurus) were the first capitosaurs with closed otic notch. On the other hand, during the Late Triassic (Carnian and Norian), the genus Cyclotosaurus displayed closed otic notch resulting in the strongest capitosaur skull during bi- and unilateral bites. Otherwise, recent studies have pointed out the biomechanical implications of the shape of the posterior skull margin, being more or less concave and providing extensive insertional area for epaxial muscles that would have assisted the cleidomastoideus during skull raising (Sulej and Majer, 2005; Maganuco et al., 2009). Epaxial muscles, which extend from the vertebral column to the cranium, inserting on the posterior region of the skull (Kathe, 1999) are part of the skull-raising system of virtually all amphibians (Lauder and Shaffer, 1993).The importance of accessory occipito-vertebral musculature on the skull-raising system should be tested in a future FEA 3D model as well as the contribution of the adductor muscles (the pterygoidei and adductor mandibulae, according to Chernin and Cruickshank, 1978) in closing the mouth. Further examination of the functional consequences of different levels of mandibular articulation and the tooth row, as the case in Cyclotosaurus versus Paracyclotosaurus (Sulej and Majer, 2005) is also required.
CONCLUSIONS Capitosaurs were generalist top-level predators of freshwater-brackish ecosystems with varied feeding capabilities. For the group as a whole, our analysis shows a trend to concentrate the stress at the circumorbital region, where more sutures were present in comparison with the rest of the skull roof, facilitating the dissipation and absorption of the stress and energy caused by biting loads. FEA revealed significant mechanical changes in the skull during the Triassic, resulting in some diversification of diet preferences within the clade. Basal taxa, such as Wetlugasaurus and Edingerella, exhibited the weakest skulls according to the modeling of bi- and unilateral biting loads. Parotosuchus exhibited one of the most strong and stable skulls and it was probably able to hunt prey of diverse sizes. In contrast, the skull of Eocyclotosaurus probably was more suited to capturing small rapid prey using lateral biting. Cyclotosaurus exhibited stronger skull with probably more capabilities to hunt larger prey, in contrast with paracyclotosaurids that showed weaker bite capabilities.
SKULL MECHANICS IN CAPITOSAURS
Our simulations suggest that the traditional hypothesis that lateral repositioning of the tabular and subsequent closure of the otic notch in some forms was driven by a need to strengthen the skull against forces generated during skull-raising (Howie, 1970) should be approached with caution. Closing of the otic notch is favored when the tabular horns are posterolaterally or laterally directed. By contrast, FEA indicates that taxa exhibiting posteriorly directed tabular horns (open otic notch) are as strong as taxa with closed otic notches, disputing previous interpretations. The cause (or causes) of the first changes in otic notch morphology remains an open question. Our results indicate that constraints apart from otic notch morphology affected the functional morphology of the group. These include elongation of the tabular horns, snout length, skull width, and orbit position and size. Previous works considered capitosaurs as ecologically analogous to ‘‘crocodylomorphs,’’ and on the base of our results we could consider capitosaurs as feeding analogues of extant crocodiles (Eusuchia: Crocodylia), only excluding gavialids and tomistonins.
ACKNOWLEDGEMENTS Authors thank Dr. J. Sebastien Steyer (MNHN, Paris), Dr. Mikhail Shishkin and Dr. Igor Novikov (PIN, Moscow), and Dr. Johannes Mu¨ller and Dr. Florian Witzmann (MfN, Berlin) for access to the collections under their care. They also thank Albert G. Selles and Soledad de Esteban-Trivigno for helpful discussions and comments and to Sonia Segura for her assistance in designing the figures. The quality of this manuscript was greatly improved with the comments of Michael Buchwitz (TU Bergakademie, Freiberg) and three anonymous reviewers.
LITERATURE CITED Benton MJ, Tverdokhlebov VP, Surkov MV. 2004. Ecosystem remodelling among vertebrates at the Permian-Triassic boundary in Russia. Nature 432:97–100. Bolt JR. 1974. Evolution and functional interpretation of some suture patterns in Palaeozoic labyrinthodont amphibians and other lower tetrapods. J Paleontol 48:434–458. Busbey AB. 1995. The structural consequences of skull flattening in crocodilians. In: Thomason JJ, editor. Functional morphology in vertebrate paleontology. Cambridge: Cambridge University Press. p 173–192. Carroll RL. 2007. The Palaeozoic ancestry of salamanders, frogs, and caecilians. Zool J Linn Soc-Lond 150:1–140. Carroll RL, Holmes R. 1980. The skull and jaw musculature as guides to the ancestry of salamanders. Zool J Linn Soc-Lond 68: 1–40. Chernin S, Cruickshank ARI. 1978. The myth of the bottom-dwelling capitosaur amphibians. S Afr J Sci 74:111–112. Cleuren J, De Vree F. 2000. Feeding in crocodilians. In: Schwenk K, editor. Feeding: Form, function, and evolution in tetrapod vertebrates. San Diego: Academic Press. p 337–358. Currey JD. 1987. The evolution of the mechanical properties of amniote bone. J Biomech 20:1035–1044. Damiani RJ. 2001a. A systematic revision and phylogenetic analysis of Triasic mastodonsauroids (Temnospondyli: Stereospondyli). Zool J Linn Soc-Lond 133:379–482. Damiani RJ. 2001b. Cranial anatomy of the giant Middle Triassic temnospondyl Cherninia megarhina and a review of feeding in mastodonsaurids. Palaeontol Afr 37:41–52.
1145
Damiani RJ. 2008. A giant skull of the temnospondyl Xenotosuchus africanus from the Middle Triassic of South Africa and its ontogenetic implications. Acta Palaeontol Pol 53:75–84. Damiani R, Schoch RR, Hellrung H, Werneburg R, Gastou S. 2009. The plagiosaurid temnospondyl Plagiosuchus pustuliferus (Amphibia: Temnospondyli) from the Middle Triassic of Germany: Anatomy and functional morphology of the skull. Zool J Linn Soc-Lond 155:348–373. Deban SM, Wake DB. 2000. Aquatic feeding in Salamanders. In: Schwenk K, editor. Feeding: Form, function and evolution in tetrapod vertebrates. New York: Academic Press. p 65–94. Defauw SL. 1989. Temnospondyl amphibians: A new perspective on the last phases in the evolution of the Labyrinthodontia. Mich. Academician 21:7–32. Doblar e M, Garcı´a JM, G omez MJ. 2004. Modelling bone tissue fracture and healing: A review. Eng Fract Mech 71: 1809–1840. Dumont ER, Grosse IR, Slater G. 2009. Requirements for comparing the performance of finite element models of biological structures. J Theor Biol 256:96–103. Erickson GM, Catanese J, Keaveny TM. 2002. Evolution of the biomechanical material properties of the femur. Anat Rec 268: 115–124. Fletcher TM, Janis CM, Rayfield EM. 2010. Finite element analysis of ungulate jaws: Can mode of digestive physiology be determined? Palaeontol Electron 13:1–15. ` , De Santisteban C. 2011a. A new capitosaur Fortuny J, Galobart A from the Middle Triassic of Spain and the relationships within the Capitosauria. Acta Palaeontol Pol 56:553–566. Fortuny J, Marc e-Nogu e J, De Esteban-Trivigno S, Gil L, Galobart ` . 2011b. Temnospondyli bite club: Ecomorphological patterns of A the most diverse group of early tetrapods. J Evol Biol 24: 2040–2054. Howie AA. 1970. A new capitosaurid labyrinthodont from East Africa. Palaeontology 13:210–253. Jenkins FAJr, Shubin NH, Gatesy SM, Warren A. 2008. Gerrothorax pulcherrimus from the Upper Triassic Fleming Fjord Formation of East Green land and a reassessment of head lifting in temnospondyl feeding. J Vertebr Paleontol 28:935–950. Kathe W. 1995. Morphology and function of the sutures in the dermal skull roof of Discosauriscus austriacus Makowsky, 1876 (Seymouriamorpha; Lower Permian of Moravia) and Onchiodon labyrinthicus Geinitz, 1861 (Temnospondyli, Lower Permian of Germany). Geobios-Lyon 28:255–261. Kathe W. 1999. Comparative morphology and functional interpretation of the sutures in the dermal skull roof of temnospondyl amphibians. Zool J Linn Soc-Lond 126:1–39. Lanyon LE, Rubin RT. 1985. Functional adaptation in skeletal structures. In: Hildebrand M, Bramble DM, Liem KF, Wake DB, editors. Functional vertebrate morphology. Harvard: Belknap Press. p 1–25. Lauder GV, Shaffer HB. 1993. Design of feeding systems in aquatic vertebrates: Major patterns and their evolutionary interpretations. In: Hanken J, Hall BK, editors. The vertebrate skull, Vol. 3: Functional and evolutionary mechanisms. Chicago: University of Chicago Press. p 113–149. Maganuco S, Steyer JS, Pasini G, Boulay M, Lorrain S, B en eteau A, Auditore M. 2009. An exquisite specimen of Edingerella madagascarensis (Temnospondyli) from the Lower Triassic of NW Madagascar: Cranial anatomy, phylogeny, and restorations. Mem Soc Ital Sci Nat Mus Civ Stor Nat Milano 36:1–72. Markey MJ, Marshall CR. 2007. Terrestrial-style feeding in a very early aquatic tetrapod is supported by evidence from experimental analysis of suture morphology. Proc Natl Acad Sci USA 104: 7134–7138. Milner AR. 1990. The radiations of temnospondyl amphibians. In: Taylor PD, Larwood GP, editors. Major evolutionary radiations. Oxford: Clarendon Press. p 321–349. Mukherjee D, Ray S, Sengupta DP. 2010. Preliminary observations on the bone microstructure, growth patterns, and life habits of some Triassic temnospondyls from India. J Vertebr Paleontol 30: 78–93.
1146
FORTUNY ET AL.
Ochev VG. 1966. [Systematics and phylogeny of capitosauroid labyrinthodonts.] Saratov: Saratov State University Press. 184 p. [in Russian]. Pawley K. 2006. The postcranial skeleton of Temnospondyls (Tetrapoda: Temnospondyli). PhD Thesis. Melbourne: La Trobe University Press. 442 p. Pierce SE, Angielczyk KD, Rayfield EJ. 2008. Patterns of morphospace occupation and mechanical performance in extant crocodilian skulls: A combined geometric morphometric and finite element modeling approach. J Morphol 269:840–864. Pierce SE, Angielczyk KD, Rayfield EJ. 2009. Shape and mechanics in thalattosuchian (Crocodylomorpha) skulls: Implications for feeding behaviour and niche partitioning. J Anat 215:555–576. Rafferty KL, Herring SW. 1999. Craniofacial sutures: Morphology, growth and in vivo masticatory strains. J Morphol 242:167–179. Rayfield EJ. 2004. Cranial mechanics and feeding in Tyrannosaurus rex. Proc Roy Soc Lond B Biol Sci 271:1451–1459. Rayfield EJ. 2005. Using finite-element analysis to investigate suture morphology: A case study using large carnivorous dinosaurs. Anat Rec 283A:349–365. Rayfield EJ. 2007. Finite element analysis in vertebrate morphology. Annu Rev Earth Planet Sci 35:541–576. Rayfield EJ, Milner AC. 2008. Establishing a framework for archosaur cranial mechanics. Paleobiology 34:494–515. Ruta M, Pisani D, Llyod GT, Benton MJ. 2007. A supertree of Temnospondyli: Cladogenetic patterns in the most species-rich group of early tetrapods. Proc Roy Soc Lond B Biol Sci 274:3087–3095. Sanchez S, Germain D, DE Ricqle`s A, Abourachid A, Goussard F, Tafforeau P. 2010. Limb-bone histology of temnospondyls: Implications for understanding the diversification of palaeoecologies and patterns of locomotion of Permo-Triassic tetrapods. J Evol Biol 23: 2076–2090. Schoch RR. 1999. Comparative osteology of Mastodonsaurus giganteus (Jaeger, 1828) from the Lettenkeuper (Longobardian) of Germany (Baden-Wu¨rttemberg, Bayern, Thu¨ringen). Stuttgarter Beitr Naturkd B 278:1–175. Schoch RR. 2000a. The origin and intrarelationships of Triassic capitosaurid amphibians. Palaeontology 43:705–727. Schoch RR. 2000b. Biogeography of Triassic capitosaur amphibians. Neues Jahrb Geol P-A 214:177–200. Schoch RR. 2008. The Capitosauria (Amphibia): Characters, phylogeny, and stratigraphy. Palaeodiversity 1:189–226. Schoch RR. 2009. Life-cycle evolution as response to diverse lake habitats in paleozoic amphibians. Evolution 63:2738–2749. Schoch RR, Milner AR. 2000. Stereospondyli. In: Wellnhofer P, editor. Encyclopedia of Paleoherpetology 3B. Munchen: Verlag Dr. Friedrich Pfeil 203 p. Schoch RR, Fastnacht M, Fichter J, Keller T. 2007. Anatomy and relationships of the Triassic temnospondyl Sclerothorax. Acta Palaeontol Pol 52:117–136. Sengupta DP, Gosh P. 1993. Morphometrics of some Triassic temnospondyls. In: Lucas SG, Morales M, editors. the Non marine Triassic. New Mexico Museum of Natural History and Science Bulletin, 3: Albuquerque, p 423–428. Sheets HD. 2003. IMP-Integrated Morphometrics Package. Buffalo, NY: Department of Physics, Canisius College.
Shishkin MA. 1987. Evolution of early amphibians (Plagiosauroidea). Akademiya Nauk SSSR Trudy Paleontologicheskogo Instituta (Trans Paleontol Inst Acad Sci USSR) 225:1–143. [in Russian] Stayton CT, Ruta M. 2006. Geometric morphometrics of the skull roof of stereospondyls (Amphibia: Temnospondyli). Palaeontology 49:307–337. Steyer JS. 2000. Ontogeny and phylogeny in temnospondyls amphibians: A new method of analysis. Zool J Linn Soc-Lond 130: 449–467. Steyer JS. 2003. A revision of the Early Triassic ‘‘Capitosaurs’’ (Stegocephali, Stereospondyli) from Madagascar, with remarks on their comparative ontogeny. J Vertebr Paleontol 23:544–555. Sulej T, Majer D. 2005. The temnospondyl amphibian Cyclotosaurus from the Upper Triassic of Poland. Palaeontology 48:157–170. Taylor MA. 1987. How tetrapods feed in water: A functional analysis by paradigm. Zool J Linn Soc-Lond 91:171–195. Warren AA. 2000. Secondarily aquatic temnospondyls of the Upper Permian and Mesozoic. In: Heatwole H, Carroll RL, editors. Amphibian Biology Vol. 4: Palaeontology. The evolutionary history of amphibians. Chipping Norton: Surrey Beatty & Sons. p 1121– 1149. Watson DMS. 1958. A new labyrinthodont (Paracyclotosaurus) from the Upper Trias of New South Wales. Bull. Brit Mus (Nat Hist) Geol 3:233–263. Watson DMS. 1962. The evolution of the labyrinthodonts. Phil Trans Roy Soc Lond B 245:219–265. Welles SP, Cosgriff J. 1965. A revision of the labyrinthodont family Capitosauridae and a description of Parotosaurus peabody, N. Sp. from the Wupatki member of the Moenkopi Formation of Northern Arizona. Los Angeles: University of California publications in Geological Sciences, Vol. 54, p 1–148. Witzmann F, Pfretzschner H-U. 2003. Larval ontogeny of Micromelerpeton credneri (Temnospondyli, Dissorophoidea). J Vertebr Paleontol 150:815–834. Witzmann F, Scholz H. 2007. Morphometric study of allometric skull growth in the temnospondyl Archegosaurus decheni from the Permian/Carboniferous of Germany. Geobios-Lyon 40: 541–554. Witzmann F, Scholz H, Ruta M. 2009. The skull ontogeny of temnospondyls: A geometric morphometrics approach. Alcheringa 33: 237–255. Witzmann F, Scholz H, Mu¨ller J, Kardjilov N. 2010. Sculpture and vascularization of dermal bones, and the implications for the physiology of basal tetrapods. Zool J Linn Soc-Lond 160:302– 340. Yates AM, Warren AA. 2000. The phylogeny of the ‘higher’ temnospondyls (Vertebrata: Choanata) and its implications for the monophyly and origins of the Stereospondyli. Zool J Linn Soc-Lond 128:77–121. Zienkiewicz OC, Zhu JZ. 1992a. The Superconvergent patch recovery and a posteriori error estimators. I The recovery technique. Int J Numer Methods Eng 33:1331–1364. Zienkiewicz OC, Zhu JZ. 1992b. The Superconvergent patch recovery and a posteriori error estimators. II Error estimates and adaptivity. Int J Numer Methods Eng 33:1365–1382.