The Face of SiamopithecusNew Geometric-Morphometric Evidence for Its Anthropoid Status.код для вставкиСкачать
THE ANATOMICAL RECORD 292:1734–1744 (2009) The Face of Siamopithecus: New Geometric-Morphometric Evidence for Its Anthropoid Status CHRISTOPH P.E. ZOLLIKOFER,1* MARCIA S. PONCE DE LEÓN,1 YAOWALAK CHAIMANEE,2 RENAUD LEBRUN,1 PAUL TAFFOREAU,3 SASIDHORN KHANSUBHAAND,2 AND JEAN-JACQUES JAEGER4 1 Anthropological Institute, University of Zürich, Zürich, Switzerland 2 Department of Mineral Resources, Paleontological Section, Bureau of Paleontology and Museum, Bangkok, Thailand 3 European Synchrotron Radiation Facility, Grenoble, France 4 Institut International Paléoprimatologie et Paléontologie Humaine, Evolution et Paléoenvironments, Université de Poitiers, Poitiers, France ABSTRACT Amphipithecids assume a key position in early primate evolution in Asia. Here we report on new maxillofacial and associated mandibular remains of Siamopithecus eocaenus, an amphipithecid primate from the Late Eocene of Krabi (Thailand) that currently represents the most complete specimen belonging to this group. We used synchrotron microtomography and techniques of virtual reconstruction to recover the three-dimensional morphology of the specimen. Geometric-morphometric analysis of the reconstructed specimen within a comparative sample of recent and fossil primates clearly associates Siamopithecus with the anthropoids. Like modern anthropoids, Siamopithecus displays a relatively short face and highly convergent and frontated orbits, the lower rim of which lies well above the alveolar plane. The cooccurrence of spatially correlated anthropoid features and classical anthropoid dental characters in one individual represents a strong argument to support the anthropoid status of Siamopithecus. It is, thus, highly unlikely that amphipithecids are specialized adapiforms exhibiting complete convergence with anthropoids. Anat Rec, C 2009 Wiley-Liss, Inc. 292:1734–1744, 2009. V Key words: primate evolution; virtual reconstruction; geometric morphometrics; synchrotron tomography; Amphipithecidae Siamopithecus eocaenus from the late Eocene lignite deposits of Krabi (Southern Thailand) was described as a new genus and species (Chaimanee et al., 1997) and, subsequently, attributed to Amphipithecidae (Jaeger et al., 1998). Based on a suite of mostly dental diagnostic characters, S. eocaenus was identified as an anthropoid of relatively large body size (Chaimanee et al., 1997; Chaimanee et al., 2000a; Egi et al., 2004). Various derived dental features suggest, on a functional level, dietary specialization (Chaimanee, 2004; Kay et al., 2004a). Siamopithecus shares its anthropoid characters with Pondaungia, Amphipithecus, and Myanmarpithecus, three late middle Eocene primate genera, from the Pondaung Formation in Myanmar (Chaimanee et al., 1997; Takai et al., 2001; Chaimanee, 2004). These SouthC 2009 WILEY-LISS, INC. V Additional Supporting Information may be found in the online version of this article. Grant sponsors: Swiss National Science Foundation, The Fyssen Foundation, The Department of Mineral Resources (Bangkok), The C.N.R.S.-T.R.F. Biodiversity project, The C.N.R.S. ‘‘Eclipse-1 & 2’’ Research Program, and The ESRF; Grant sponsor: Swiss NFS; Grant numbers: N 205321-102024/1and 205320-109303/1. *Correspondence to: Christoph P.E. Zollikofer, Anthropological Institute, University of Zürich Winterthurerstrasse 190, CH8057 Zürich, Switzerland. Fax: þ41 44 635 6886. E-mail: zolli@ aim.uzh.ch Received 12 May 2009; Accepted 15 June 2009 DOI 10.1002/ar.20998 Published online 28 August 2009 in Wiley InterScience (www. interscience.wiley.com). SIAMOPITHECUS RECONSTRUCTION 1735 Fig. 1. S. eocaenus specimen TF 7624-7625 from the Late Eocene locality of Krabi, Thailand. A: Associated maxillary and mandibular remains before physical preparation (site photograph). B: Occlusal view of the maxillae following preparation. Scale bar is 1 cm. east Asian primates exhibit an array of peculiar synapomorphies, which warrant their inclusion in Amphipithecidae (Jaeger et al., 1998; Chaimanee et al., 2000a; Beard, 2002; Chaimanee, 2004; Jaeger et al., 2004; Takai and Shigehara, 2004). Consensus has now been reached concerning the monophyly of that group (Kay et al., 2004b; Jaeger and Marivaux, 2005; Marivaux et al., 2005; Seiffert et al., 2005). Jaeger et al. (2004) argue that the two largest taxa, Amphipithecus and Pondaungia, display no significant morphological difference other than size (but see Takai and Shigehara, 2004, for an opposing view), such that they might represent males and females of a single, sexually dimorphic species. Pronounced sexual dimorphism, which among primates is only known in anthropoids (Kelley and Qinghua, 1991; Simons and Plavcan, 1999; Simons et al., 2007), would thus represent an additional anthropoid character of amphipithecids. In an alternative interpretation, amphipithecids are associated with the adapiforms (Kay et al., 2004a), specifically with the notharctids (Ciochon and Holroyd, 1994; Ciochon and Gunnell, 2004), and are seen as an early evolutionary convergence with anthropoids (Gunnell and Miller, 2001; Kirk and Simons, 2001). Following this ‘‘complete convergence’’ hypothesis, the hypothesized specialization to hard diets (Chaimanee et al., 1997; Jaeger et al., 1998; Chaimanee, 2004) would represent a major evolutionary constraint shaping the entire amphipithecid cranial morphology (Kay et al., 2004a). However, an analysis of dental enamel microstructure by means of synchrotron microtomography (SR-lCT) suggested that, at least in Siamopithecus, a diet based on hard food is improbable (Tafforeau, 2004): Siamopithecus has thin, radial enamel, which is not in accordance with high-pressure resistance implied by hard food items. The ‘‘complete convergence’’ hypothesis has other difficulties. For example, amphipithecids share no single derived dentognathic character with notharctids (Kay et al., 2004a). Another disputed point concerns a postcranial partial skeleton (humerus, calcaneus, and ulna fragment; NMMP 20), which has tentatively been associated with the largest Pondaung amphipithecids and which is described as exhibiting notharctid affinities (Ciochon et al., 2001). However, a talus (NMMP 39) from the Pondaung Formation displays all the diagnostic derived characters of anthropoids (Marivaux et al., 2003). Because the talus is considered to represent a critical postcranial element for primate phylogeny (Gebo et al., 2000), the evidence from the Pondaung talus is in stark contrast with that from the other postcranial elements. The recent discovery of a diversified sivaladapid community in the Pondaung Formation (Beard et al., 2007; Marivaux et al., 2008) renders more probable the hypothesis that the NMMP 20 postcranial remains belong to a large and dentally still undocumented sivaladapid. Moreover, a frontal bone fragment (NMMP 19), which was ascribed to a small-sized Pondaung amphipithecid (Gunnell et al., 2002; Shigehara et al., 2002; Takai et al., 2003) and was considered to provide evidence for absence of postorbital closure, is most likely of nonmammalian origin, as evinced by a detailed comparative reanalysis of its anatomy (Beard et al., 2005). Here, we examine the ‘‘anthropoid’’ versus ‘‘complete convergence’’ hypotheses of amphipithecid origins in the light of previously undescribed fossil evidence bearing on Siamopithecus. The 1996 excavations in the Bang Mark pit of the Krabi lignite mine yielded a new specimen of S. eocaenus, consisting of a right mandible (TF 7625; Chaimanee et al., 2000a) preserved in anatomical occlusion with midfacial remains (TF 7624). The latter comprise both maxillae including the hard palate and, on the left side, parts of the orbital rim formed by the zygomatic bone (Fig. 1). The upper dentition is represented on both sides by P3-M3 and the alveoli of P2 and C. The premaxillary bone and incisors are lacking. The specimen directly confirms the association of upper and lower molars proposed earlier on the basis of isolated gnathic elements (Chaimanee et al., 1997; Ducrocq, 1999) and permits a first view of amphipithecid facial morphology. This study has two aims: the first is to present the results of a virtual reconstruction of the face of Siamopithecus and to identify key features of amphipithecid facial morphology. The second is to perform a comparative geometric-morphometric analysis of its restored morphology. Geometric-morphometric methods are independent of classical methods of craniodental character 1736 ZOLLIKOFER ET AL. analysis, such that the phenetic analyses presented here yield valuable complementary data to test taxonomic and phylogenetic hypotheses. MATERIALS AND METHODS State of Preservation of the Specimen Like many other fossils recovered from coal deposits, the TF 7624/7625 specimen underwent distortion through fragmentation and taphonomic compression. Because physical preparation would involve unnecessary risk and correction of plastic deformation is impractical, we performed a virtual reconstruction. Given the small size and high degree of mineralization of the TF 7624/ 7625 remains, we used SR-lCT (Tafforeau et al., 2006) with a monochromatic beam at 65 keV to acquire digital volume data at an isotropic voxel size of 45.71 lm. When compared with conventional lCT, the high-energy, highflux, monochromatic beam of SR-lCT has the advantage of yielding cross-sectional images with high spatial and contrast resolution and free of beam hardening artifacts (Tafforeau et al., 2006; Fig. 2). Semiautomated image segmentation procedures were applied to separate virtual fossil parts along major cracks and to diagnose the specimen’s external and internal state of preservation. The dental arcades of the right and left maxillae remained undisturbed postmortem, with the exception of the left M3 (as evinced by mirror-image matching of the two sides). The preserved right mandibular corpus underwent substantial lateral compression, which resulted in fracturing and mediolateral flattening. During this process, fragments were crushed and driven apart in a vertical direction, such that the original height of the mandibular corpus (except for the symphyseal region) cannot be reconstructed with confidence. Virtual Reconstruction Fig. 2. Synchrotron microtomographic parasagittal cross-section of the right maxilla of S. eocaenus TF 7625. Scale bar is 1cm. The isolated parts served as a basis for four virtual reconstructions, which were carried out independently by four team members (CZ, MPL, RL, PT), following the general principles outlined in Zollikofer and Ponce de Fig. 3. Stages of virtual reconstruction of Siamopithecus TF 7624/25 (reconstruction protocol 1; see text). A: establishment of dental occlusion (right lateral view). B: Correction of deformation of the mandibular corpus (basal view; transparent: distorted original morphology). C, D: correction of plastic deformation in the right palatal area (transparent: distorted original morphology) and reposition of the left zygomaxillary fragment. Scale bar is 5 cm. SIAMOPITHECUS RECONSTRUCTION Fig. 4. Virtual reconstruction of S. eocaenus TF 7624-7625 following protocol 1 (see text). A: right lateral view; B: frontal view; C, D: superior views; E, F: inferior views. Colors indicate electronically isolated parts; mirror-image completions are transparent; scale bar is 5 cm. 1737 1738 ZOLLIKOFER ET AL. TABLE 1. Comparative sample Family Strepsirrhines Cheirogaleidae Indriidae Lemuridae Lepilemuridae Galagidae Lorisidae Adapidaey Notharctidaey Archaeolemuridaey Haplorrhines Tarsiidae Aotidae Atelidae Cebidae Pitheciidae Cercopithecidae Hylobatidae Pliopithecidaey Omomyidaey y TABLE 2. Landmarks N 18 8 9 9 16 24 7 1 6 4 6 11 17 10 41 4 1 1 Extinct. León (2005). Reconstructions 1–3 started with reestablishment of dental occlusion between the well-preserved right maxilla and the isolated mandibular teeth (Fig. 3). Subsequently, the isolated fragments of the mandibular corpus and ramus were adapted using dental positions as a guide. The left and right maxillary halves exhibit anatomical contact along the palate. The noticeable plastic deformation of the right side of the palate was corrected with reference to the better-preserved left side, such that the maxillae could be oriented relative to the midsagittal plane of the skull. Positioning the maxillae with the reconstructed right mandible and its mirror image in dental occlusion showed that the mandibular symphysis was crushed mediolaterally during fossilization (Fig. 4). To recover the midfacial architecture of the specimen, maxillary fragments on the right side were repositioned through comparison with mirror-imaged matching regions on the left side. The position and orientation of the orbital rim fragment was evaluated by adapting a mirror image of the isolated left zygomatic bone to the well-preserved right zygomatic process of the maxilla and subsequent adjustment of the corresponding fragmentary region on the left side. During reorientation and relocation of the zygomatic bone, the mandibular coronoid process served as an additional positional clue. This structure must fit into the temporal fossa, the anterolateral extent of which is constrained by the zygomatic bone. In comparison with its displaced in situ position, the reconstructed zygomatic fragment is elevated relative to the alveolar plane and assumes a more anteroposterior orientation. The fourth virtual reconstruction (Tafforeau 2004) also started with the reconstruction of the mandible, but it used undistorted mandibles of Pondaungia (NMMP 24) and of Amphipithecus as templates to reconstruct the dental arch of Siamopithecus. This reconstruction is based on the hypothesis that the specific ‘‘parabolic’’ shape of the mandibular dental arch is a synapomorphy of amphipithecids. The vertical curvature (i.e., the shape of the occlusal plane) was reconstructed according to the morphology of the relatively undistorted Siamopithecus mandibular frag- Number Maxillofacial 1 2, 3 4, 5 6, 7 8, 9 10, 11 12, 13 14, 15 Mandibular 16 17 18, 19 20, 21 22, 23 24, 25 26, 27 Landmark definition Prosthion Frontomalare orbitale Orbitale Point between fmo and o Buccalmost point on canine Buccalmost point on P2 or P3 Buccalmost point on M1 Buccalmost point on M3 Infradentale Gnathion Buccalmost point on Buccalmost point on Buccalmost point on Buccalmost point on Foramen mentale canine P2 or P3 M1 M3 ment TF3634. The reconstructed right mandible was then complemented with its mirror image, and the mediolateral inclination of the hemimandibles was adjusted according to the NMMP24 mandible of Pondaungia. The resulting mandibular reconstruction exhibits a narrower symphyseal region compared with reconstructions 1–3. The upper jaws were reconstructed with the better-preserved right maxilla and its mirror-image; these were placed in anatomical position relative to each other and put in occlusion with the mandibular dentition. Finally, the left zygomatic bone was adapted to the maxillary reconstruction and mirror imaged to the right side. The consensus of reconstructions 1–3 is shown in Fig. 4 (individual reconstructions are shown in Supporting Information Figs. S1 and S2). The spatial consistency of dental occlusion in the reconstructive variants was checked with 3D-hardcopies produced by means of 3Dprinting technology. Reconstructions 1–3 converged with reconstruction 4 in an important feature: the shape of the mandibular dental arcade (which resulted from dental occlusion with the undistorted right maxilla) is similar to that of the undistorted Pondaungia mandible, which served as a starting point for reconstruction 4. Reconstruction 4 resulted in some degree of anisognathy (i.e., different widths of upper and lower dental arcs): masticatory movements inferred from this reconstruction—as well as the radial enamel structure revealed by SR-lCT—suggest a folivorous diet in Siamopithecus (Tafforeau, 2004). Morphometric Analysis We used geometric-morphometric methods to assess the phenetic position of the reconstructed Siamopithecus morphology (as represented by the four reconstructive variants) within a comparative sample of N ¼ 194 extant and fossil prosimian and anthropoid primate skulls (Table 1; details in Supporting Information Table S1). Craniomandibular form was quantified with 15 maxillofacial and 12 mandibular anatomical landmarks, whose location could be determined reliably on the virtual reconstructions of Siamopithecus (Table 2). Patterns of shape variation in the sample were analyzed using principal components (PC) analysis of shape SIAMOPITHECUS RECONSTRUCTION 1739 Fig. 5. Principal components analysis of craniomandibular and cranial shape. A: craniomandibular shape variability along the first three PCs. B: corresponding virtual morphologies at extreme values of PC1 (0.25; 0.21), PC2 (0.21; 0.25), and PC3 (0.15; 0.09). C: cranial shape variability along the first three PCs. D: corresponding virtual morphologies at extreme values of PC1 (0.22; 0.21), PC2 (0.22; 0.28), and PC3 (0.11; 0.22). Arrows in A and C show axes of sizerelated shape variation in strepsirrhines and anthropoids. In B and D, crania of Lepilemur ruficaudatus AIMZ11054 (negative pole of PC1) and Saimiri sciureus AIMZ9159 (all other cases) are used to represent shape transformations. Grey/black symbols: strepsirrhines/haplorrhines. Horizontal rectangles: lemuriforms; vertical rectangles: lorisiforms; diamonds: Archaeolemur; open triangles: Adapinae; filled triangle: Notharctus; stars: catarrhines; circles: platyrrhines; z: Tarsius; þ: Microchoerus; square: Pliopithecus; X: four reconstructive variants of Siamopithecus. in Linearized Procrustes space (Dryden and Mardia, 1998), after shape variation was constrained to bilateral symmetry (Zollikofer and Ponce de León, 2002). Analysis and visualization of patterns of shape variation were performed with the interactive geometric morphometrics software package MorphoTools (Specht, 2007; Specht 1740 ZOLLIKOFER ET AL. of reference. The orbital plane (O) is determined by three landmarks on the zygomatic portion of the orbital rim (frontomalare orbitale, orbitale, and a point between these landmarks). The cranial midplane M is evaluated by averaging the positions of bilateral landmark pairs and calculating the plane through these points. The alveolar plane A is defined as containing the landmarks on maxillary canines and third molars (see Table 2 for landmark definitions). Using a standard definition of vector geometry, the orientation of planes O, M, and A is given by the corresponding normal vectors nO, nM, nA (Fig. 7B). These vectors permit measurement of orbital convergence (angle between nO and nM) and frontation (angle between nO and nA). RESULTS Fig. 6. Phenetic trees depicting morphological affinities of Siamopithecus with extant and extinct primate families. A: UPGMA phenetic tree computed for the skull. B: UPGMA phenetic tree computed for the cranium. Siamopithecus branches within anthropoids. Grey: prosimians (Strepsirrhini þ Tarsiiformes); black: anthropoids. et al., 2007; Lebrun, 2008). Size-related shape variation (size allometry) was quantified by evaluating allometric shape vectors, which were obtained by multivariate regression of PC scores on log centroid size. Because strepsirrhine and anthropoid primates differ widely in patterns of craniomandibular shape variation, groupspecific allometric shape vectors were evaluated. To quantify morphological affinities between Siamopithecus and the comparative primate sample, Procrustes shape distances were computed between Siamopithecus and extant and fossil primate taxa (for extant taxa, familyspecific mean shapes were used). By using PHYLIP (Felsenstein, 1989), the resulting distance matrix was represented as a phenetic tree. To evaluate the Siamopithecus reconstruction within the comparative framework of earlier studies using conventional distance- and angle-based morphometrics, we derived various linear and angular measurements from the 3D landmark data. The definitions of orbital frontation and convergence proposed by Ross (1995) cannot be applied here directly, because the neurobasicranial and medial orbital morphology of Siamopithecus is missing. To calculate analogous variables, we used three planes Graphing the first three PCs, which account for 74% of the total shape variability in the sample, shows clear separation between prosimian (Tarsius þ strepsirrhines) and anthropoid morphologies (Fig. 5A,B). This demonstrates that a small but reliable set of maxillofacial and mandibular landmarks provides sufficient analytical sensitivity to resolve taxonomically relevant morphological differences in the sample. Similar analyses were performed on maxillofacial morphology alone (Fig. 5C,D), permitting inclusion of various fossil crania without associated mandibles, such as those of early primates (Adapiformes and Omomyidae) and of subfossil lemurs (Archaeolemur). Inclusion of Archaeolemur in the comparative sample is of interest, because overall craniofacial form in this genus exhibits a high degree of convergence with anthropoid primates (Forsyth-Major, 1896). In all analyses, Siamopithecus groups with the anthropoids (Figs. 4–6). As an additional check of consistency of the virtual reconstruction, we tested whether Siamopithecus could be reconstructed to fit the morphology of known fossil and extant nonanthropoid primates. To this end, the Siamopithecus landmark configuration was transformed into the closest possible hypothetical prosimian configuration, and the virtual fossil parts were accommodated accordingly. All reconstructions toward strepsirrhine morphologies led to anatomically impossible configurations, involving overlap and/or disruption of anatomical continuity between neighboring fragments. Among the fossil specimens, adapiform primates (Magnadapis, Leptadapis, Adapis, and Notharctus), whose nonanthropoid status is undisputed, are more closely associated with prosimians, whereas Pliopithecus groups with the anthropoids. As expected, the archaeolemurs present a morphology that is closer to that of anthropoids than other strepsirrhines, but all of them are clearly distinct from anthropoids in shape space (Fig. 5C). Furthermore, Siamopithecus exhibits smaller shape distances to most anthropoid families than to extant or extinct prosimian families (Fig. 6). Major morphological contrasts within the sample can be visualized by shape transformation of an average primate face towards extreme shape values of the data scatter along each PC (Fig. 5B,D). PC1, which accounts for most of the differences between anthropoids and prosimians, reveals contrasts between the short-snouted, comparatively narrow anthropoid face with a high mandibular corpus and orbital cavities that are elevated above the alveolar plane and exhibit a high degree of SIAMOPITHECUS RECONSTRUCTION Fig. 7. Relative orbital dimensions and orbital orientation. A: relative biorbital width (fmo: absolute biorbital width; M3C: distance between buccal alveolar borders of M3 and C; M1: distance between buccal alveolar borders of M1). B: orbital frontation versus orbital convergence (in degrees). Frontation (f) is the angle between nA and nO; 1741 convergence (c) is the angle between nO and nM (see text). Open/ closed circles: diurnal/nocturnal anthropoids; open/closed squares: diurnal/nocturnal strepsirrhines. Diamonds: Archaeolemur. X: reconstructive variants of Siamopithecus; Y: Pliopithecus; Z: Tarsius; open triangles: Adapinae; filled triangle: Notharctus; þ: Microchoerus. 1742 ZOLLIKOFER ET AL. TABLE 3. Morphological characters of Siamopithecus and amphipithecidsa Anthropoid characters of Siamopithecus eocaenus shared with large Amphipithecidae from Pondaung Cingulum-derived hypocone Almost continuous crista obliqua and short trigone’s basin Well-developed hypoparacrista Upper premolars unwaisted in occlusal view Reduction or absence of labial cingula on molars Deep horizontal branch of lower jaw Strong bunodonty with molar talonid displaying nearly the same elevation as trigonid Reduced, single-rooted P/2 P/3-P/4 exodaenodont and obliquely oriented P/4 with well-developed metaconid cusp Straight cristid oblique Presence of X-facet on lower molars Vertical symphysis Synapomorphies of Amphipithecidae Crest linking protocone to hypocone on upper molars Very high horizontal branch of the mandible Short dental rows indicating short muzzles Parabolic tooth rows in occlusal view Convex upper occlusal surface and corresponding concave lower occlusal surface Wrinkled enamel Absence of paraconid Waisted lower molar outline in occlusal view Reduced to absent hypoconulid Slanted buccal and lingual upper molar walls M/3 surface smaller than M/2 Entoconid reduced and distally displaced Cingulids reduced to absent Plesiomorphic characters of Amphipithecidae Presence of P2/2 Unfused symphysis a Jaeger and Marivaux, 2005, Science 310:244–245; Seiffert et al., 2005, Science 310:300–304. TABLE 4. Linear and angular dimensions of the virtual reconstruction of Siamopithecus Method of measurement Dimensiona Bi-frontomolare orbitale (fmo) Between alveolar midpoints Between alveolar midpoints Between lingual alveolar borders of M1 Between buccal alveolar borders of M1 Between buccal alveolar borders of M1 Infradentale-gnathion See text See text (fmo-fmo)/(M1-M1) (fmo-fmo)/(M3-C) 52.9 mm 29.7 mm 33.9 mm 20.6 mm 39.1 mm 35.6 mm 25.7 mm 81.6 degrees 72.1 degrees 1.35 mm 1.78 mm Variable Biorbital width C-M3 (maxilla) C-M3 (mandible) Palate width (at M1) Maxillary width at M1 M1-M1 buccal width (mandible) Symphyseal height Frontation Convergence Biorbital width rel. to bimaxillary width Biorbital width rel. to maxillary length a Mean values of the four reconstructive variants. convergence, as opposed to the long, low prosimian face with a low mandibular ramus, and with orbits at the level of the alveolar plane and exhibiting a low degree of convergence. PC2 mostly accounts for allometric shape variation in the sample, which influences the degree of prognathism and of orbital frontation, while PC3 expresses a pattern of variation similar to that of PC2, but independent of size. Although the orbital morphology of Siamopithecus is only partially preserved, it is possible to derive various functionally relevant measurements. Siamopithecus had orbits of moderate size compared with gnathic dimen- sions (Fig. 7A), and it clearly falls within the anthropoid range of variation of frontation and convergence. DISCUSSION In the context of the current discussion of an anthropoid versus strepsirrhine affiliation of the amphipithecids, the reconstructed face of Siamopithecus provides new data on previously unknown aspects of amphipithecid morphology, which are relevant for taxonomic and phyletic inferences. First and foremost, the shape of the face of Siamopithecus, as quantified by a configuration 1743 SIAMOPITHECUS RECONSTRUCTION of 3D anatomical landmarks, clearly falls within the variation displayed by extant and fossil anthropoids and outside the variation displayed by extant and fossil prosimians. A suite of correlated characters can be identified in these analyses, which all group Siamopithecus with anthropoids as follows: 1. The position of Siamopithecus in shape space (Figs. 5, 6) suggests that its overall facial morphology is close to that of a generalized anthropoid. 2. The lower margin of the orbit is well above the alveolar plane, as in most anthropoids, and unlike in prosimians. 3. The high degree of orbital frontation and convergence inferred from the reconstruction (Fig. 4B and Table 3) place Siamopithecus within the variation displayed by anthropoid primates and outside the range of variation of prosimians. 4. Biorbital distance (Fig. 7A and Table 4) suggests that the orbits were of moderate relative size and in the range of diurnal anthropoids (Kay and Kirk, 2000). 5. The mandibular morphology of Siamopithecus is reminiscent of that of the undeformed Burmese amphipithecid specimens (Chaimanee et al., 2000b; Jaeger et al., 2004). The mandibular dental arch displays a parabolic shape between P3 and M3, the ramus is high, and the canines were large (judging from the preserved parts of the right lower canine and from the reconstructed border of the alveolar socket of the right upper canine). The face of Siamopithecus is, thus, best described as that of a generalized anthropoid. Moreover, the combination of facial features seen in Siamopithecus (short maxilla, high degree of orbital convergence and frontation, and relatively small orbits well above the alveolar plane) is also characteristic for moderate-sized extant anthropoid primates, such as cebids, atelids, langurs, and small cercopithecines (e.g., Miopithecus). However, Siamopithecus differs from langurs and atelids by exhibiting larger canines and a more robust zygomatic process. The virtual reconstruction presented here also permits a first comparative assessment of midfacial variability within amphipithecids. The only other known amphipithecid midface—that of Pondaungia cotteri (NMMP 18; Shigehara et al., 2002)—is less complete than that of S. eocaenus, but several features can be compared directly. In both specimens, the facial surface of the maxilla bulges laterally and, in inferior view, overarches the posterior dental arcade. The height of the maxillary body, as estimated from the distance between the alveolar margin and the orbital rim, suggests a large maxillary sinus in both specimens. A robust zygomatic process is also a feature found in both amphipithecid specimens, but the root of the zygomatic arch is located more mesially in Siamopithecus (above M1) than in Pondaungia (above M2), indicating a shorter face in the former species. Combining the new morphometric data with classical morphological evidence provides additional support for the hypothesis that Siamopithecus, and its related Burmese amphipithecids are anthropoid primates. A list of dentognathic characters of Siamopithecus and the large amphipithecids compiled from previous publications (Jaeger and Marivaux, 2005; Seiffert et al., 2005) is provided in Table 3. This character complex is indicative of an anthropoid affiliation of the amphipithecids. Although some of these ‘‘anthropoid’’ characters also occur in several prosimians (Ciochon and Gunnell, 2004), the large array of concurrent anthropoid dentognathic features shared by all amphipithecids is so far documented only for undisputed anthropoid primates. Amphipithecids also exhibit an array of plesiomorphic features characteristic of primitive primates (Table 3). Interestingly, however, some of these features are also present in several Fayum late Eocene anthropoids (Beard, 2002; Rasmussen, 2002) and may be related to the greater geological age of these fossils compared with lower Oligocene Fayum crown anthropoids displaying more derived character states. Our analyses also confirm that craniofacial convergence of Archaeolemur with anthropoids is only superficial. The archaeolemurid specimens examined here all have anteroposteriorly short faces; however, geometric-morphometric analysis clearly groups them with strepsirrhines. A recent virtual reconstruction of the subfossil archaeolemurid Hadropithecus also displays a cranial morphology reminiscent of anthropoids (Ryan et al., 2008), and it remains to be examined whether this peculiar morphology matches the pattern of morphological convergence of archaeolemurids toward anthropoid morphologies revealed by our geometric-morphometric analyses. Overall, the ‘‘complete convergence’’ hypothesis of amphipithecids toward anthropoids is less parsimonious than the ‘‘anthropoid origins’’ hypothesis, and the postulated association of amphipithecids with prosimian primates seems to be an effect of incomplete fossil evidence and of misattribution of some fragmentary cranial and postcranial specimens. Additional comparative fossil evidence is, thus, required to better understand the phyletic relationships and the dietary and locomotor specializations of this early Asian anthropoid group. ACKNOWLEDGMENTS We greatly acknowledge Dr. med. K. Geissmann’s support with medical CT. We thank the staff of beamlines ID19 and ID17 (European Synchrotron Radiation Facility, ESRF), and Peter Wyss (EMPA) for help with microtomography. Special thanks to Matthias Specht for collaborative implementation of MorphoTools. We thank Suzanne Jiquel and Monique Vianey-Liaud (I.S.E.M.), Edmée Ladier (Musée d’Histoire Naturelle de Montauban), Jacques Cuisin, Marc Godinot and Pascal Tassy (Museum National d’Histoire Naturelle de Paris) for access to primate specimens. Literature cited Beard KC. 2002. Basal anthropoids. In: Hartwig WC, editor. The primate fossil record. Cambridge: Cambridge University Press. p 133–149. Beard KC, Jaeger J-J, Chaimanee Y, Rossie JB, Soe AN, Tun ST, Marivaux L, Marandat B. 2005. Taxonomic status of purported primate frontal bones from the Eocene Pondaung Formation of Myanmar. J Hum Evol 49:468–481. Beard KC, Marivaux L, Tun ST, Soe AN, Chaimanee Y, Htoon W, Marandat B, Aung HH, Jaeger J-J. 2007. New sivaladapid primates from the Eocene Pondaung Formation of Myanmar and the anthropoid status of Amphipithecidae. Bull Carnegie Mus Nat Hist 39:67–76. Chaimanee Y. 2004. Siamopithecus eocaenus, anthropoid primate from the late Eocene of Krabi, Thailand. In: Ross CF, Kay RF, 1744 ZOLLIKOFER ET AL. editors. Anthropoid origins: new vision. New York: Kluwer Academic/Plenum Publishers. p 341–368. Chaimanee Y, Khansubhaa S, Jaeger J-J. 2000a. A new lower jaw of Siamopithecus eocaenus from the Late Eocene of Thailand. C R Acad Sci Ser 323:235–241. Chaimanee Y, Suteethorn V, Jaeger J-J, Ducrocq S. 1997. A new Late Eocene anthropoid primate from Thailand. Nature 385: 429–431. Chaimanee Y, Thein UT, Ducrocq S, Soe AN, Benammi M, Tun T, Lwin T, Wai S, Jaeger J-J. 2000b. A lower jaw of Pondaungia cotteri from the Late Middle Eocene Pondaung Formation (Myanmar) confirms its anthropoid status. Proc Natl Acad Sci USA 97: 4102–4105. Ciochon RL, Gingerich PD, Gunnell GF, Simons EL. 2001. Primate postcrania from the late middle Eocene of Myanmar. Proc Natl Acad Sci USA 98:7672–7677. Ciochon RL, Gunnell GF. 2004. Eocene large-bodied primates of Myanmar and Thailand: morphological considerations and phylogenetic affinities. In: Ciochon RL, Holroyd PA. 1994. The Asian origin of Anthropoidea revisited. In: Fleagle J, Kay RF, editors. Anthropoid origins. New York: Plenum Press. p 143–162. Dryden IL, Mardia KV. 1998. Statistical shape analysis. New York: Wiley. Ducrocq S. 1999. Siamopithecus eocaenus, a late Eocene anthropoid primate from Thailand: its contribution to the evolution of anthropoids in Southeast Asia. J Hum Evol 36:613–635. Egi N, Tun ST, Takai M, Shigehara N, Tsubamoto T. 2004. Geographical and body size distributions of the Pondaung primates with a comment on the taxonomic assignment of NMMP 20, postcranium of an amphipithecid. Anthropol Sci 112:67–74. Felsenstein J. 1989. PHYLIP-Phylogeny Inference Package (Version 3.2). Cladistics 5:164–166. Forsyth-Major CI. 1896. Preliminary notice on fossil monkeys from Madagascar. Geol Mag 3:433–436. Gebo DL, Dagosto M, Beard KC, Qi T, Wang J. 2000. The oldest known anthropoid postcranial fossils and the early evolution of higher primates. Nature 404:276–278. Gunnell GF, Ciochon RL, Gingerich PD, Holroyd PA. 2002. New assessment of Pondaungia and Amphipithecus (Primates) from the late middle Eocene of Myanmar, with a comment on ‘Amphipithecidae’. Contrib Mus Paleontol Univ Mich 30:337–372. Gunnell GF, Miller ER. 2001. Origin of anthropoidea: dental evidence and recognition of early anthropoids in the fossil record, with comments on the Asian anthropoid radiation. Am J Phys Anthropol 114:177–191. Jaeger J-J, Chaimanee Y, Tafforeau P, Ducrocq S, Soe AN, Marivaux L, Sudre J, Thura S, Htoon W, Marandat B. 2004. Systematics and paleobiology of the anthropoid primate Pondaungia from the late middle Eocene of Myanmar. C R Palevol 3:241–253. Jaeger J-J, Marivaux L. 2005. Shaking the earliest branches of anthropoid primate evolution. Science 310:244–245. Jaeger J-J, Soe AN, Aung AK, Benammi M, Chaimanee Y, Ducrocq R-M, Tun T, Thein UT, Ducrocq S. 1998. New Myanmar middle Eocene anthropoids. An Asian origin for catarrhines? C R Acad Sci Ser 3 321:953–959. Kay RF, Kirk EC. 2000. Ostological evidence for the evolution of activity pattern and visual acuity in primates. Am J Phys Anthropol 113:235–262. Kay RF, Schmitt D, Vinyard CJ, Perry JM, Shigehara N, Takai M, Egi N. 2004a. The paleobiology of Amphipithecidae, South Asian late Eocene primates. J Hum Evol 46:3–25. Kay RF, Williams BA, Ross CF, Takai M, Shigehara N. 2004b. Anthropoid origins: a phylogenetic analysis. In: Ross CF, Kay RF, editors. Anthropoid origins: new visions. New York: Kluwer Academic/Plenum Publishers. p 91–135. Kelley J, Qinghua X. 1991. Extreme sexual dimorphism in a Miocene hominoid. Nature 352:151–153. Kirk EC, Simons EL. 2001. Diet of fossil primates from the Fayum Depression of Egypt: a quantitative analysis of molar shearing. J Hum Evol 40:203–229. Lebrun R. 2008. Evolution and development of the strepsirrhine primate skull. PhD Thesis. Zurich, Germany: University Montpellier II and University of Zürich. Marivaux L, Antoine PO, Baqri SRH, Benammi M, Chaimanee Y, Crochet J-Y, Franceschi D, Iqbal N, Jaeger J-J, Métais G, Welcome J-L. 2005. Anthropoid primates from the Oligocene of Pakistan (Bugti Hills): data on early anthropoid evolution and biogeography. Proc Natl Acad Sci USA 102:8436–8441. Marivaux L, Beard KC, Chaimanee Y, Jaeger J-J, Marandat B, Soe AN, Tun ST, Aung HH, Htoon W. 2008. Anatomy of the bony pelvis of a relatively large-bodied strepsirrhine primate from the late middle Eocene Pondaung Formation (central Myanmar). J Hum Evol 54:391–204. Marivaux L, Chaimanee Y, Ducrocq S, Marandat B, Sudre J, Soe AN, Tun ST, Htoon W, Jaeger J-J. 2003. The anthropoid status of a primate from the late middle Eocene Pondaung Formation (Central Myanmar): tarsal evidence. Proc Natl Acad Sci USA 100:13173–13178. Rasmussen DT. 2002. Early catarrhines of the African Eocene and Oligocene. In: Hartwig WC, editor. The primate fossil record. Cambridge: Cambridge University Press. p 203–220. Ross CF. 1995. Allometric and functional influences on primate orbit orientation and the origins of the Anthropoidea. J Hum Evol 29:201–227. RossCF, Kay RF, editors. Anthropoids origins: new visions. New York: Kluwer Academic/Plenum Publishers. p 249–282. Ryan TM, Burnay DA, Godfrey LR, Göhlich UB, Jungers WL, Ramilisonina, Walker A, Weber GW. 2008. A reconstruction of the Vienna skull of Hadropithecus stenognathus. Proc Natl Acad Sci USA 105:10699–10702. Seiffert ER, Simons EL, Clyde WC, Rossie JB, Attia Y, Bown TM, Chatrath P, Mathison ME. 2005. Basal anthropoids from Egypt and the antiquity of Africa’s higher primate radiation. Science 310:300–304. Shigehara N, Takai M, Kay RF, Aung AK, Soe AN, Tun ST, Tsubamoto T, Thein T. 2002. The upper dentition and face of Pondaungia cotteri from central Myanmar. J Hum Evol 43:143–166. Simons EL, Plavcan JM. 1999. Canine sexual dimorphism in Egyptian Eocene anthropoid primates: Catopithecus and Proteopithecus. Proc Natl Acad Sci USA 98:2559–2562. Simons EL, Seiffert ER, Ryan TM, Attia Y. 2007. A remarkable female cranium of the early Oligocene anthropoid Aegyptopithecus zeuxis (Catarrhini, Propliopithecidae). Proc Natl Acad Sci USA 104:8731–8736. Specht M. 2007. Spherical surface parameterization and its application to geometric morphometric analysis of the braincase. PhD Thesis. Zurich, Germany: University of Zurich. Specht M, Lebrun R, Zollikofer CPE. 2007. Visualizing shape transformation between chimpanzee and human braincases. Vis Comp 23:743–751. Tafforeau P. 2004. Phylogenetic and functional aspects of tooth enamel microstructure and three-dimensional structure of modern and fossil primate molars: contributions of X-ray synchrotron microtomography. PhD Thesis. Hérault, France; University of Montpellier II. Tafforeau P, Boistel R, Boller E, Bravin A, Brunet M, Chaimanee Y, Cloetens P, Feist M, Hoszowska J, Jaeger J-J, Kay RF, Lazzari V, Marivaux L, Nel A, Nemoz C, Thibault X, Vignaud P, Zabler S. 2006. Applications of X-ray synchrotron microtomography for non-destructive 3D studies of paleontological specimens. Appl Phys A 83:195–202. Takai M, Shigehara N. 2004. The Pondaung Primates, enigmatic ‘‘possible anthropoids’’ from the latest middle Eocene, central Myanmar. In: Ross CF, Kay RF, editors. Anthropoid origins: new visions. New York: Kluwer Academic/Plenum Publishers. p 283–321. Takai M, Shigehara N, Aung AK, Tun ST, Soe AN, Tsubamoto T, Thein T. 2001. A new anthropoid from the latest middle Eocene of Pondaung, central Myanmar. J Hum Evol 40: 393–409. Takai M, Shigehara N, Tsubamoto T. 2003. Endocranial cast and morphology of the olfactory bulb of Amphipithecus mogaungensis (latest middle Eocene of Myanmar). Primates 44:137–144. Zollikofer CPE, Ponce de León MS. 2002. Visualizing patterns of craniofacial shape variation in Homo sapiens. Proc R Soc B 269: 801–807. Zollikofer CPE, Ponce de León MS. 2005. Virtual reconstruction. A primer in computer-assisted paleontology and biomedicine. Hoboken, New Jersey: Wiley.