Homology of the Jaw Muscles in Lizards and Snakes A Solution from a Comparative Gnathostome Approach

Transcription

1 THE ANATOMICAL RECORD 297: (2014) Homology of the Jaw Muscles in Lizards and Snakes A Solution from a Comparative Gnathostome Approach PETER JOHNSTON* Department of Anatomy with Radiology, University of Auckland, Private Bag 92019, Auckland, New Zealand Abstract Homology or shared evolutionary origin of jaw adductor muscles in lizards and snakes has been difficult to establish, although snakes clearly arose within the lizard radiation. Lizards typically have temporal adductors layered lateral to medial, and in snakes the muscles are arranged in a rostral to caudal pattern. Recent work has suggested that the jaw adductor group in gnathostomes is arranged as a folded sheet; when this theory is applied to snakes, homology with lizard morphology can be seen. This conclusion revisits the work of S.B. McDowell, J Herpetol 1986; 20: , who proposed that homology involves identity of m. levator anguli oris and the loss of m. adductor mandibulae externus profundus, at least in advanced (colubroid) snakes. Here I advance the folded sheet hypothesis across the whole snake tree using new and literature data, and provide a solution to this homology problem. Anat Rec, 297: , VC 2014 Wiley Periodicals, Inc. Key words: snakes; lizards; myology; comparative morphology Snakes evolved within the branches of the lizard phylogenetic tree, although exactly where in the tree is a matter of debate (Lee, 2009; Losos et al., 2012). The evolutionary trajectory in snakes has been toward the macrostomatan condition, allowing a wide gape to swallow whole prey and relying on constriction or venom to subdue prey, rather than biting. This trend has resulted in the reduction of the mass of the adductor muscle and loss of aponeurotic structures in favor of increased fibre length. The result of this evolution has been that jaw adductor muscles of macrostomatan snakes appear to have a different configuration to those of lizards and Sphenodon: the muscle bellies of snake adductors appear to be layered rostral to caudal, and those of the other extant lepidosaurs are layered lateral to medial (Haas, 1973; McDowell, 1986; Zaher, 1994). Basal alethinophidian lineages (see cladogram, Fig. 1) are less modified from the typical lizard condition (Rieppel, 1980), and scolecophidians ( blindsnakes ) have different highly apomorphic feeding mechanisms that are associated with specialised anatomy (Kley, 2001). Establishing homology among the adductor muscles and arriving at a consistent nomenclature are unresolved issues in lepidosaur morphology. This is demonstrated in Figure 2, which shows superficial views of the musculature in the iguanid lizard Ctenosaura pectinata and in Python sebae, with terminology according to various authors applied. It can be seen that with the conventional terminology (Lakjer, 1926; Haas, 1973; Rieppel, 1980) there are three components to the external adductor in Python. A key component of this homology problem is the identification of the most superficial layer of the typical lepidosaurian pattern (Daza et al., 2011), the levator anguli oris muscles that attach to the rictal plate (Mundplatte) at the corner of the mouth. McDowell (1986) addressed the homology problem by suggesting that in colubroid snakes, the most rostral external adductor component corresponds to the lizard m. levator anguli oris (LAO), which has become modified in Abbreviations used: LAO 5 levator anguli oris; MAME 5 m. adductor mandibulae externus. *Correspondence to: Peter Johnston; Department of Anatomy with Radiology, University of Auckland, Private Bag 92019, Auckland, New Zealand. psjohnston@clear.net.nz Received 2 January 2013; Accepted 27 May DOI /ar Published online 31 January 2014 in Wiley Online Library (wileyonlinelibrary.com). VC 2014 WILEY PERIODICALS, INC.

2 SNAKE JAW ADDUCTORS 575 Fig. 1. Family level cladogram of extant snakes, modified from Wiens et al. (2012). Asterisk indicates the families dissected. Inset photographs include some representatives of the species and genera studied. Photographs credits: Damian Hyde, ReptileConnexion, Australia (Antaresia children, Morelia bredli), Bryan Hughes (Crotalus atrox) Eric B. Holt (Nerodia fasciata), Palot and Radhakrishnan (2010) (Hydrophis platurus). various ways to accommodate the venom gland, which is itself a development of the rictal plate. In this model, parts or all of m. levator anguli oris need not insert on the rictal plate, but attach instead to the venom gland or to the mandible by fascial or direct insertion. McDowell also suggested that one component of the m. adductor mandibulae externus (MAME), m. adductor mandibulae externus profundus, is absent in snakes; with these two proposals, the homology between lizard and snake adductors can be resolved. Subsequent authors have not agreed with or extended McDowell s model. In a recent account of the homology of the jaw adductors among Lissamphibia, Johnston (2011) proposed that the adductor muscle group could be visualized as a folded sheet, and that this represented a better basis for understanding homology than the traditional system based on position of trigeminal nerve branches within the muscle group. This folded sheet morphology (Fig. 3) was examined across various gnathostome groups and found to be a consistent pattern, and to be capable of explaining muscle homologies from chondrichthyan fish to mammals. Insertion of the jaw adductor group onto the mandible is in an inverted U shape, folded around the m. adductor mandibulae posterior, which is regarded as being of different homology to the rest of the adductor group. In this system, components of the muscle group are best defined by their position of insertion into the U. Snakes were not examined in that work; the present account looks at snake jaw adductors in light of the folded sheet system, and finds this to be fully consistent with the homology of McDowell (1986), and that this homology can be extended across all snake radiations to resolve the lizard-snake jaw adductor problem. The current account is based on new data on representative snakes, details of jaw muscles in a wide gnathostome group (Johnston, 2011) and in lizards and Sphenodon (Daza et al., 2011), and literature reports of snake morphology. Sphenodon is assumed to be close to the plesiomorphic lepidosaur condition for jaw adductors, but for purposes of comparison with snakes (Fig. 2), the iguanid lizard Ctenosaura pectinata is chosen, as absence of the ventral temporal arch makes illustration clearer.

3 576 JOHNSTON Fig. 2. Comparison of lizard and snake jaw adductor muscles. A, B: Ctenosaura pectinata (after Oelrich, 1956), lateral view; D, E: Ctenosaura pectinata : dorsal view; C: Python sebae (after Frazetta, 1966), lateral view, F: Python sebae, dorsal view. Homologous muscles, according to the present work and McDowell (1986) are indicated in A and C. Identifications of muscles by other authors are given in boxes. Literature data across a large number of snakes will be summarized together with a brief historical account of snake jaw muscle identity and nomenclature. New data will then be presented, and the snake jaw adductor muscles will be discussed as a folded sheet morphology that is compatible with the situation in lizards and other gnathostomes. NOMENCLATURE OF JAW ADDUCTORS IN SNAKES The account of Lakjer (1926) has been influential in identification of components of the snake jaw adductor group. His system followed that of Luther (1914), who assigned major divisions in the musculature according to trigeminal nerve branches. Other schemes of muscle

4 SNAKE JAW ADDUCTORS 577 focussed on aponeurotic systems in the main body of the adductor. Zaher noted that the literal definition of the m. levator anguli oris resulted in state reversals in the phylogeny of alethinophidians: presence of LAO in anilioids is interpreted as the primitive state, followed by loss of LAO in booids, with regain of the muscle in colubroids. Zaher found this puzzling, and commented on the frailty of definition of LAO. The differences among these authors are obvious, and are compounded by Haas (1973) and Rieppel (1980) using traditional terminology while recognizing homologies that are at variance with these terms. Homology as used here refers to primary homology hypotheses sensu De Pinna (1991), based on anatomical (including developmental) similarity and congruence, as opposed to secondary homology hypotheses, which are deduced following cladistic analysis. Fig. 3. Proposed morphology of tetrapod jaw muscle adductors, here demonstrated in the amphicoelous frog Ascaphus truei: the muscle(s) take the form of a sheet folded around the m. adductor mandibulae posterior, which is a separate entity. M. levator anguli oris, where present, is also separate to this pattern (see Fig. 7). Modified from Johnston (2011). Terminology is given as typically used in amphibians (in italics, underlined) (Haas, 2001; Johnston, 2011) (m. levator or adductor mandibulae internus-longus-externus-lateralis) and in amniotes (plain font), see Table 1. nomenclature have been proposed by Edgeworth (1935) and Diogo and Abdala (2010). A summary of equivalent muscle identifications for snakes and lizards in a wider tetrapod context is given in Table 1, and Table 2 gives a detailed comparison of homologies of the superficial muscles critical to the argument presented here. It is m. levator anguli oris and m. adductor mandibulae externus, the part of the adductor muscle rostrolateral to the mandibular (V 3 ) nerve, that give rise to difficulties with homology. Lakjer (1926) and most subsequent authors have interpreted m. levator anguli oris in a literal sense, as only applicable to muscle attaching to the rictal plate; Underwood (1967) and McDowell (1986) have allowed a more liberal interpretation, to include topologically equivalent muscles with insertions to mandible and venom gland. Lakjer (1926) was somewhat unsure what to call the most caudal and ventral section of the adductor group, and hesitated between externus medialis and externus profundus ; he chose the latter (Fig. 2), as it corresponded to a muscle he had thus named in birds. This designation has been followed by most authors, although its position in the most superficial layers of the adductor is in conflict with this nomenclature. Lakjer (1926) relied on the Bodenaponeurose, the large fibrous tendon of insertion present in lizards and Sphenodon, to separate the components of the adductor externus, but macrostomatan snakes do not have this structure, leading to some of the difficulties in homology and nomenclature. In addition, the loss of the dorsal temporal arch and post-temporal fenestra in snakes makes interpretation of the temporal heads of the adductor difficult, but Rieppel (1980) in a study of primitive snakes Anilius, Cylindrophis and Uropeltis provided clarification, at least in this group. Zaher (1994) gave considerable detail on the adductor muscles, and JAW ADDUCTORS IN SNAKE PHYLOGENY A cladogram is presented in Figure 1, with taxa investigated here indicated on the right side. Topology of the snake tree remains controversial among recent accounts (Conrad, 2008; Burbrink and Crother, 2011; Scanlon and Lee, 2011; Gauthier et al., 2012; Wiens et al., 2012); the general pattern of Scolecophidia as basal radiation(s), booids and pythons in a central position and colubroids in the upper branches of the tree is conserved, but many disagreements on individual branches are unresolved. Scolecophidia The highly apomorphic feeding mechanisms of leptotyphlopids and typhlopids (Kley, 2001) are reflected in their cranial muscles, but the configuration of jaw adductors in leptotyphlopids is in keeping with the pattern found in Alethinophidia, with a prominent LAO, and rostral to caudal layered adductor bellies. Haas (1930, 1973) published different interpretations of the external layers of the adductor group, recognising a more extensive LAO in his later work. In typhlopids the typical pattern is much distorted by the presence of an autapomorphic m. retractor maxillae, which covers the temporal region of the skull and lies deep to a reduced adductor group. The identity of typhlopid adductor muscles has not been addressed since studies by Haas (1973) and Iordansky (1997); both interpret MAME superficialis (MAMES) as lying deep to both MAME medialis and profundus, and this finding requires further examination. Anomalepidids have been studied only by Haas (1964, 1968) in serial sections; he admitted to difficulty achieving satisfactory reconstructions from this data, and some of his diagrams are difficult to understand. Alethinophidia: Basal Lineages The retention of some lizard features in Anilius and Cylindrophis is pointed out by Haas (1973), Rieppel (1980), and Zaher (1994): origin of MAMES from a fascial band which may represent the dorsal temporal arch, presence of a Bodenaponeurose and the medial and dorsal extent of the adductor externus. Haas (1973) recognised that fibres corresponding to MAME profundus of lizards are present in these taxa, but avoided renaming this part of muscle to reduce confusion. There are two

5 578 JOHNSTON TABLE 1. Systems of homology and nomenclature of jaw muscles This work 1 McDowell (1986) Lakjer (1926) 1 Haas (1973) Edgeworth (1935): all Lizards Snakes Lizards Snakes tetrapods See Table 2 Add. mand. externus M. pseudotemporalis Add. mand. medius M. pterygoideus M. pterygoideus M. pterygoideus M. pterygoideus M. retractor Add. mand. maxillae internus (typhlopids) Constrictor dorsalis: mm. levator bulbi, protractor pterygoidei, levator pterygoidei Diogo and Abdala (2010): bony fish, tetrapods A2 A3 0 A3 00 PTM M. adductor mandibulae posterior PVM Constrictor dorsalis: mm. retractor pterygoidei, retractor vomeris; protractor pterygoidei, levator pterygoidei add. mand., m. adductor mandibulae. Constrictor dorsalis: mm. levator bulbi, protractor pterygoidei, levator pterygoidei, retractor maxillae (typhlopids) Constrictor dorsalis: mm. retractor pterygoidei, retractor vomeris, protractor pterygoidei, levator pterygoidei Constrictor internus dorsalis Dorsal mandibular muscles TABLE 2. Adductor externus terminology and homology according to previous authors [Modified from Zaher (1994)] Lizards, sensu Rieppel (1980) Snakes, sensu Haas (1973) Snakes, sensu Rieppel (1980) Snakes, sensu McDowell (1986) Snakes, sensu Zaher (1994) Levator anguli oris, 1a Levator anguli oris Levator anguli oris Levator anguli oris Levator anguli oris Adductor externus Adductor externus Superficialis proper1fibres Levator anguli oris Adductor externus superficialis 1b superficialis of temporal head of superficialis1 superficialis 1b (posteroventral part) medialis (pinnate part) medialis (quadrate head) Adductor externus profundus 3a Adductor externus profundus 3b13c Adductor externus medialis Adductor externus profundus medialis Adductor externus medialis Adductor externus profundus Ext, externus; med, medialis; sup, superficialis. superficialis profundus, superficial fibres Fibres of temporal head and lateral fibres of add ext. sup.1med Deep fibres of add. ext. profundus Deep fibres of add. ext. profundus Deep fibres of adductor ext. medialis superficialis medialis superficialis superficialis lost temporalis medialis pars posterior temporalis1medialis pars posterior medialis pars posterior medialis pars posterior profundus bundles of muscle in the rostral border of the muscle group; these are interpreted by the above authors as one bundle of LAO (ames1a), inserting onto the rictal plate, and one belly of MAMES (ames1b), which inserts onto the mandible. Booid and Pythons Similar patterns are seen in booids (Zaher, 1994) and pythons (Frazzetta, 1966) (Fig. 2): a rostral muscle (MAME superficialis of Lakjer [1926]) inserts via a tendon over the most caudal section (MAME profundus of Lakjer), with a medialis section between: this pattern represents the 3 externi morphology recognised by Haas (1973), with the three muscle bellies stacked rostro-caudal. Colubroids Kochva (1962) detailed many viperids, in which part of the MAME profundus of Lakjer has become modified to arise from the venom gland, and may be termed m. compressor glandulae, although the same name is given by other authors to modifications of the LAO. Considerable reduction of the medialis section of the muscle is noted. Other important accounts of colubroid muscles were published by Liem et al. (1971), Kardong (1973), and Cundall (1983). McDowell (1986) and Zaher (1994) documented the many variations of attachment of the adductor muscle group to the venom gland. McDowell noted that the two bundles of muscle present insert on the venom gland and mandible respectively, but one or other may lie superficial; he proposed that the rictal plates are a

6 SNAKE JAW ADDUCTORS 579 specific anatomical territory, to which the venom gland and its attached muscles belong; hence his conclusion that both these sections of muscle should be interpreted as LAO. Underwood (1967) made a similar point: both parts of the muscle may insert on the mandible via aponeurosis, directly or onto the venom gland; in some taxa only one larger muscle is present, and he considered these muscles should be regarded as a unit. The development of the jaw adductors has been studied in only two taxa, and is thus of limited contribution to homology considerations. Detailed accounts of Vipera palestinae by Kochva (1963) and Natrix natrix by Rieppel (1988) both found that the muscle group attaching to the venom gland and corner of the mouth separated early from the main adductor block, which would be consistent with their identity as LAO, although these authors did not discuss this. Edgeworth (1935) outlined muscle development in Tropidonotus, presumably also Natrix natrix, and a lateral and rostral muscle is seen to separate from the adductor externus (his Figs ), similar to the above authors results. MATERIALS AND METHODS Specimens examined were: Leptotyphlops humilis cahuilae CAS80440; Typhlops lumbricalis AIM LH no number; Ramphotyphlops braminus AIM LH2282; Boa constrictor JVM H503, 504, 505; Antaresia childreni JVM H650, 651; Morelia spilotes JVM H723, 724; Nerodia fasciata JVM H862, 863; Hydrophis platurus (formerly Pelamis platurus) AIM LH1961; Alsophis elegans AIM LH331; Crotalus atrox AIM LH1276; Sphenodon punctatus OUA AA, A-Z. Abbreviations: CAS: California Academy of Sciences; AIM: Auckland Museum; JVM: P. Johnston, personal collection; OUA: Otago University, Dunedin, Anatomy Department. Leptotyphlops humilis and Typhlops lumbricalis were examined by microct scanning with Skyscan 1172 machines, after soft tissue staining with phosphotungstic acid (Metscher, 2009). 3D reconstructions were made with Amira (Visage Imaging Inc.). Other taxa were dissected under an SZ-600 dissecting microscope, with occasional use of Lugol s iodine staining (Bock and Shear, 1972) to define muscle fibre direction. Dissections were photographed through the microscope with a Casio Exilim HR EX-ZR100 camera, and the images were used for the basis of drawings in Corel Graphics Suite X4. Three-dimensional models to demonstrate adductor muscle morphology as a folded sheet were made in Blender 2.49b. Abbreviations follow the convention that the muscles as identified in the current study are abbreviated in upper case, and identifications of other authors in lower case italic. LAO: m. levator anguli oris; MAME: m. adductor mandibulae externus; MAMES: MAME superficialis; MAMEM: MAME medialis; MAMEP: MAME profundus. Osteological terminology follows Cundall and Irish (2008). RESULTS The findings presented are confined to the superficial mandibular adductors, with additional features from deeper layers given only when differences in morphology or interpretation from published accounts are to be noted. Leptotyphlops humilis Two bodies of LAO are present, and insert on the rictal plate (Fig. 4A,B). The more rostral LAO inserts on the broad rictal plate, and the more caudal muscle (adductor externus 2 of Haas [1930]) inserts via a tendon that branches out within the extensive glandular tissue at the angle of the mouth; some of these tendinous bands reach the mandible. Three bellies of MAME deep are seen to these, and are here interpreted as MAME superficialis (profundus of Haas [1930]) and two bellies of MAME medialis. No equivalent of MAMEP is present, in that there is no muscle originating from the occipital part of the cranium and inserting deep to the superficialis and medialis sections. Typhlops lumbricalis and Ramphotyphops braminus A wide body of muscle considered here as MAME superficialis1medialis (profundus of Haas [1930] and Iordansky [1997]) arises from a fascial band that is attached to the dorsal process of the quadrate and overlies m. retractor maxillae (Fig. 4C, D, E). This superficial muscle inserts on the compound bone and coronoid. No MAME profundus is recognised. Deep to these a vertically oriented muscle is pierced by the maxillary nerve (Fig. 4D), and is here interpreted as m. pseudotemporalis. M. retractor maxillae is a large muscle arising from the convexity of the cranium as far as the occiput, and converging as it passes deep to the above muscles to insert on the maxilla (Fig. 4E). Two sections of m. pterygoideus are found, the more rostral of these arising from the maxilla in two tendinous bands, and the caudal part, which arises from the pterygoid. Boa constrictor MAME superficialis arises from the quadrate and converges on an insertion on the compound bone. MAME medialis is separate and arises from the dorsal, lateral rim of the parietal to insert on both compound bone and coronoid (Fig. 5A). LAO arises from postorbital bone and adjacent parietal, and inserts with a flat tendon inserts onto the fascia covering MAME superficialis. Antaresia childreni and Morelia spilotes MAME superficialis and medialis are present as a continuous sheet and converge on an insertion on the compound bone (superficialis) and coronoid (medialis) (Fig. 5B). LAO arises from the postorbital bone and adjacent parietal, lies superficial to MAME and inserts by a short flat tendon into the fascia over MAME superficialis. Insertions of the adductor muscle group onto the mandible are shown in Figure 5C. Nerodia fasciata MAME superficialis and medialis are almost continuous, arising from quadrate and parietal, and insert on compound bone and its coronoid process (Fig. 6A). LAO is a single muscle belly that arises from postorbital and

7 580 JOHNSTON Fig. 4. Jaw muscles in scolecophidian snakes. A, Leptotyphlops humilis, superficial exposure; B, Leptotyphlops humilis, after removal of the caudal element of m. levator anguli oris; C, Ramphotyphlops braminus, superficial exposure; D, Ramphotyphlops braminus, after removal of superficial adductors; E, Ramphotyphlops braminus, after removal of mandible, showing insertion of m. retractor maxillae and rostral origin of m. pterygoideus from maxilla (shaded grey). adjacent parietal, is attached to the caudal aspect of the venom gland directly and to the rictal plate by a short tendon, and inserts both with a flat tendon over MAME superficialis and with a bundle of fibres comprising about 30% of the muscle that inserts directly into the fibres of MAME superficialis and medialis. Alsophis elegans MAME superficialis and medialis are partly separated, and insert in a similar fashion to Nerodia (Fig. 6B). LAO is present in two sections; the more rostral arises from postorbital and is firmly attached to the dorsal side of the labial gland apparatus and ends in a flat tendon over MAMES. The more caudal LAO muscle arises from the parietal and is closely attached to the medial face of the labial glands on its way to insert on the compound bone, immediately rostral to the insertion of MAME medialis. Hydrophis platurus MAME superficialis and medialis are separate; MAME medialis is a relatively small muscle with insertion confined to the coronoid process of the compound bone (Fig. 6C and insert). LAO is a complex muscle with two components: a superficial part arises from the parietal and inserts into the tough covering of the venom gland, and a deep part arises mainly from postorbital and attaches firmly to the deep aspect of the venom gland before looping around the angle of the mouth to insert on the compound bone just rostral to MAME superficialis. Crotalus atrox The lateral view of the adductor muscle is dominated by the very large m. compressor glandulae, which becomes continuous toward its insertion onto the compound bone and its coronoid process with MAME superficialis, which arises from the quadrate (Fig 6D, E). MAME medialis is reduced. LAO is reduced and present deep to the large venom gland, arising from postorbital and adjacent parietal, and inserting with a flat tendon into the fascia over MAME superficialis. A stout ligament suspends the venom gland from the parietal, but its position lateral to LAO tends to rule out the possibility of this ligament being the homologue of any part of

8 SNAKE JAW ADDUCTORS 581 Fig. 5. A: Boa constrictor, jaw muscles, superficial exposure; B: Antaresia children, jaw muscles, superficial exposure; C: Antaresia childreni, dorsal view of left mandible with muscle insertions; D: Sphenodon punctatus, dorsal view of left mandible with muscle insertions, and outline of insertion of the jaw muscles as an inverted U pattern around the m. adductor posterior. the adductor muscle group. Deep to these structures lies the small MAME medialis with dorso-ventral fibre direction and inserting onto the coronoid process; the maxillary nerve V 2 separates this muscle from the equally small m. pseudotemporalis, which inserts immediately medial to MAME medialis. Sphenodon punctatus Findings are as presented in Daza et al. (2011); all components of the lepidosaur adductor pattern are present, as in Ctenosaura pectinata (Oelrich, 1956) (Fig. 2). The pattern of insertion of the jaw muscles onto the mandible is shown in Figure 5D, which also indicates the inverted U pattern of insertion mentioned above. DISCUSSION The muscle identified in the present account as LAO is regarded as a component of MAME superficialis by all other authors apart from McDowell (1986), except where insertion is clearly onto the rictal plate. The reasons for considering this to be LAO in the present work are the same as those given by McDowell (1986): (1) superficial

9 582 JOHNSTON Fig. 6. Jaw muscles in colubroid snakes. A: Nerodia fasciata; B: Alsophis elegans; C: Hydrophis platurus; D: Crotalus atrox, superficial exposure; E: Crotalus atrox, after removal of m. compressor glandulae. position, (2) a lack of continuity of muscle belly and fibre direction with other superficial layers of the adductor complex, (3) a close association with the rictal gland, labial glands or venom gland, and (4) a range of relations between parts of the muscle, where present. Additionally, the folded sheet model of Johnston (2011) can be applied if LAO is considered to be a separate rostral muscle group and not part of the folded sheet. Comparison of Figure 5C,D shows the U-shaped pattern of insertion in Antaresia and Sphenodon, and demonstrates the proposed homology of the adductor externus superficialis of lizards and Sphenodon with the adductor externus profundus of most authors on snakes (see Table 2). Further evidence for the separate nature of the LAO can be deduced from the embryological studies of Kochva (1963) and Rieppel (1988), with the early separation of the rostral muscle block that gives rise to LAO. A threedimensional model of the folded sheet with deletion of the MAME profundus section is given in Figure 7. In the unmodified amniote state (Fig. 7A), a gap is present in the sheet between m. pterygoideus and the other component of the deep adductor group, m. pseudotemporalis superficialis. The deletion of the MAME profundus (Fig. 7B, C) leaves a further gap in the sheet, and it is proposed that the absence of caudal and dorsal fibres alters the sheet to a flatter configuration, without the typical folding of other tetrapods. Implicit in the folded sheet morphology is the secondary importance of nerve trunks in defining muscle identity. A number of examples of amphibian and reptilian taxa with muscle morphology that is not consistent with nerve placement are given by Haas (2001), Holliday and Witmer (2007) and Johnston (2011). In addition, Liem et al. (1971) found m. pseudotemporalis deep to V 2 in Azemiops. A new example is given here: the m. pseudotemporalis (as identified here) in Typhlops and Ramphotyphlops. In the typhlopids, previous authors (Haas, 1930; Iordansky, 1997) have identified two muscles in these positions, separated by the V 2 nerve, but this leads to an anomalous situation of the MAME superficialis lying deep to other parts of this muscle. The present explanation (m. pseudotemporalis pierced by the V 2 ) avoids this, and retains the concept of muscle topology of a discontinuous folded sheet. The folded sheet theory emphasizes the overall shape of the adductor group rather than specifically named sections of the muscle, but the latter are important in attempting to establish homology among various amniote groups and will be discussed below. Another observation on homologies in typhlopids can be made here, regarding the unique m. retractor maxillae: the insertion of m. retractor maxillae in close proximity to the origin of the rostral bellies of m. pterygoideus (Fig. 4E) supports the suggestion of Haas (1973) that m. retractor maxillae is derived from m. pterygoideus, in contrast to Haas (1930) and Lakjer (1926), who considered it to have evolved from the constrictor internus dorsalis group. Although previous authors, with the exception of McDowell (1986), have followed the original designation of MAME profundus as the most lateral and caudal part of the muscle (see Table 2), Haas (1973) and Zaher (1994) have noted problems with this. Both have concluded that this part of the muscle is better regarded as MAME medialis, the muscle that in lizards inserts on the lateral face of the Bodenaponeurose, but Haas (1973) desisted from renaming it in the interest of consistency with other literature. The solution adopted here, MAME superficialis, is consistent with the situation in Sphenodon and lizards (Daza et al., 2011), turtles (Jones et al., 2012), crocodilians, and birds (Holliday and Witmer, 2007) in which a muscle inserts over a broad area on the lateral face of the mandible this represents the most

10 SNAKE JAW ADDUCTORS 583 Fig. 7. Left lateral view of folded sheet morphology of jaw adductors in tetrapods, modified in snakes by loss of m. adductor mandibulae profundus. Color gradients indicate transitions between component muscles. A: basic amniote pattern, with terminology as used in reptilians. A gap is present in the sheet between mm. pterygoideus and pseudotemporalis. B: hypothetical condition with loss of profundus section. C: condition in macrostomatan snakes, with reconfiguration as a flatter sheet following loss of profundus. lateral, caudal end of the folded sheet morphology. In addition, the finding of continuity or near-continuity of the MAME superficialis and MAME medialis, for example in Antaresia (Fig. 5B), Nerodia (Fig. 6A), and Alsophis (Fig. 6B) supports this view. It has been suggested that the loss of the posttemporal fenestra in snakes results in the loss of part (Haas, 1973) or the whole (McDowell, 1986) of the lizard MAME profundus. This is a very interesting suggestion, as the posttemporal fenestra is bounded by differing configurations of parietal, squamosal and supraoccipital bones both within lizards (Evans, 2008) and in reptilians in general (Romer, 1956); also, McDowell (2008) has recently rekindled an old debate about the identity of the temporal bones in snakes. This situation invites the concept of cryptic anatomical boundaries introduced by Koentges and Lumsden (1996) (chick) and Matsuoka et al. (2005) (mouse), who demonstrated highly specific territories of origin and insertion of cranial muscles, directed by neural crest and not conforming to boundaries of individual bones. The folded sheet theory includes a specific inverted U shaped insertion on the mandible (Johnston, 2011) that is consistent with such boundaries, but the association of the origins of specific muscle bellies with anatomical territories has only been made in the above models. Rieppel (1988) noted that basal lineages of alethinophidian snakes retain the Bodenaponeurose, and that the identification of muscle components in these taxa would be a critical test of the McDowell (1986) hypothesis, which Rieppel thought may not be applicable to these snakes. McDowell formulated his hypothesis specifically for colubroid snakes but did generalize the findings to all alethinophidians in his discussion. It is important, therefore, to examine the detailed findings of Rieppel (1980) in Anilius, Cylindrophis and Uropeltis, and of Zaher (1994) in Anilius and Xenopeltis, in light of the reiteration of McDowell s theory as the folded sheet model proposed here. Rieppel (1980) proposed composite homologies for a number of muscle bodies (see Table 2). A simpler conclusion, consistent with McDowell (1986) and the current proposal, can be drawn: LAO is present in two components in these taxa: in both Anilius and Cylindrophis a rostral muscle inserts on the rictal plate (lao of Rieppel and Zaher) and a second rostral belly (part of ames 1b of Reippel, aes of Zaher) inserts by a flat tendon into the fascia of MAME superficialis (amep of Rieppel, aem2 of Zaher), similar to the LAO insertion in pythons and booids (Fig. 4). In Uropeltis, two LAO muscles insert into the rictal plate (lao and ames 1b of Rieppel). In Xenopeltis, three components of LAO are present: two oblique rostral bellies associated with the rictal gland (lao of Zaher) and a third more caudally placed muscle inserting via a flat tendon as in booids (aes of Zaher). The most caudal and lateral belly of the adductor proper

11 584 JOHNSTON is MAMES, as in other snakes (Fig. 2C, F) (amep of Rieppel and aem2 of Zaher), and the remainder of the muscle, attached to the Bodenapoeurose, is best considered on topographical grounds as a fusion of MAMEM and MAMEP of the lizard model (amem of Rieppel, aem 1 of Zaher). Zaher (1994) emphasized aponeuroses as indicators of homologies within the adductor muscles, which restricted his ability to comment on homologies in macrostomatans that do not have aponeuroses. Another correlate of the folded sheet Bauplan is that aponeuroses, like nerves, may be better seen as servants rather than masters of an underlying pattern. Zaher (1994) pointed out a lack of topological criteria to define a muscle as LAO, and thus the difficulty in accepting McDowell s hypothesis, although McDowell did go to some length to emphasize the association of LAO with the rictal plate and the rictal and venom glands. The proposal advanced here, LAO as separate from the folded adductor sheet, offers a topological explanation. Can the muscle anatomy as interpreted here bring any new light to the deepening controversy over the position of snakes in the squamate tree (Gauthier et al., 2012; Losos et al., 2012; Wiens et al., 2012)? Like Rieppel (1980) in his interpretation, I cannot find any greater similarity of a snake pattern to any particular lizard or amphisbaenian group apart from to varanids, in particular Lanthanotus borneensis (Haas, 1973; Rieppel, 1980), which has a head shape similar to that of many snakes. The weakly supported clade of limbless squamates recovered by Gauthier et al. (2012) is not strengthened by the jaw muscle findings: snakes, amphisbaenians (Rieppel, 1979) and dibamids (Haas, 1973) can all be seen to be derived from a plesiomorphic pattern (Sphenodon or Ctenosaura, Fig. 2) but are not particularly similar to one another. Further investigation of jaw muscles in anomalepidids may reveal features useful in consideration of the monophyly or otherwise of scolecophidian snakes. The findings presented here can be placed in a hypothesis for the evolution of the jaw muscles in macrostomatan snakes, with adaptation for wide mouth opening rather than powerful closing: increase in fibre length by loss of aponeuroses and modification of m. levator anguli oris to act both on the venom gland and partly or entirely as an adductor. The rostral territory of origin of LAO may offer better mechanical advantage than the other components of the adductor group on the widely opened jaw. Muscular control of the angle of the mouth is added to or replaced by the various morphologies of the m. neurocostomandibularis complex (Langebartel, 1968), an apomorphy of snakes. The greatly developed venom gland in Viperidae results in reduction of m. levator anguli oris, but the viperid m. compressor glandulae, a modification of the superficial adductor (Fig. 6D), may be acting as a rostrally arising jaw adductor as well as acting on venom expulsion, as evidenced by the strong suspensory ligament that prevents ventral displacement of the gland. ACKNOWLEDGMENTS I am indebted to the following for help with loan and acquisition of specimens: Brian Gill, Auckland Museum; Mark Hutchinson and Carolyn Kovach, South Australian Museum; Sharon Moore; Juan D. Daza, Villanova University; Bob Drewes and Jens Vindum, California Academy of Sciences. Michael Byrne (Auckland University) and Andrew McNaughton (Otago University) kindly assisted with the microct scans. I am grateful to David Cundall and Rui Diogo for insightful reviews and to Juan D. Daza for editorial assistance. LITERATURE CITED Bock W, Shear C A staining method for gross dissection of vertebrate muscles. Anatomischer Anzeiger 130: Burbrink FT, Crother BI Evolution and taxonomy of snakes. In: Aldridge RD, Sever DM, editors. Reproductive biology and phylogeny of snakes. Enfield, New Hampshire: Science Publishers Inc. p Conrad J Phylogeny and systematics of Squamata (Reptilia). Bull Am Mus Nat Hist 310: Cundall D Activity of the head muscles during feeding by snakes: a comparative approach. Am Zool 23: Cundall D, Irish F The snake skull. In: Gans C, Gaunt AS, Adler K, editors. Biology of the Reptilia, vol 20 (The Skull of Lepidosauria). Ithaca, New York: Society for the Study of Amphibians and Reptiles. p Daza JD, Diogo R, Johnston P, Abdala V Jaw adductor muscles across lepidosaurs: a reappraisal. Anat Rec 294: De Pinna MGG Concepts and tests of homology in the cladistic paradigm. Cladistics 7: Diogo R, Abdala V Muscles of vertebrates: comparative anatomy, evolution, homologies and development. Enfield: Science Publishers. Edgeworth FH The cranial muscles of vertebrates. Cambridge: The University Press. Evans SE The skull of lizards and tuatara. In: Gans C, Gaunt AS, Adler K, editors. Biology of the Reptilia, vol. 20 (The Skull of Lepidosauria). Ithaca, New York: Society for the Study of Reptiles and Amphibians. p Frazzetta T Studies on the morphology and function of the skull in the Boidae (Serpentes). Part 2 Morphology and function of the jaw apparatus in Python sebae and Python molurus. J Morphol 118: Gauthier J, Kearney M, Maisano J, Rieppel O, Behlke A Assembling the squamate tree of life: perspectives from the phenotype and the fossil record. Bull Peabody Mus Nat Hist 53: Haas G Uber das Kopfskelett und die Kaumuskulatur der Typhlopiden und Gauconiiden. Zoologische Jahrb ucher, Abteilung f ur Anatomie und Ontogenie der Tiere 52:1 94. Haas G Anatomical observations on the head of Liotyphlops albirostris (Typhlopidae, Ophidia). Acta Zool (Stockh) 45:1 62. Haas G Anatomical observations on the head of Anomalepis aspinosus (Typhlopidae, Ophidia). Acta Zool (Stockh) 49: Haas G Muscles of the jaws and associated structures in the Rhynchocephalia and Squamata. In: Gans C, Parsons T, editors. Biology of the Reptilia, Volume 4. London: Academic Press. p Haas A Mandibular arch musculature of anuran tadpoles, with comments on homologies of amphibian jaw muscles. J Morphol 247:1 33. Holliday C, Witmer L Archosaur adductor chamber evolution: integration of musculoskeletal and topographical criteria in jaw muscle homology. J Morphol 268: Iordansky NN Jaw and feeding mechanics of Typhlops (Ophidia: Typhlopidae): a reconsideration. Russ J Herpet 4: Johnston P Cranial muslces of the anurans Leiopelma hochstetteri and Ascaphus truei and the homologies of the mandibular adductors in Lissamphibia and other gnathostomes. J Morphol 272:

12 SNAKE JAW ADDUCTORS 585 Jones M, Werneburg I, Curtis N, Penrose R, O Higgins P, Fagan M, Evans S The head and neck anatomy of sea turtles (Cryptodira: Chelonioidea) and skull shape in Testudines. PLOS One 7: e Kardong KV Lateral jaw and throat musculature of the cottonmouth snake Agkistrodon piscivorus. Gegenbaur s Morphologisches Jahrbuch 119: Kley N Prey transport mechanisms in blindsnakes and the evolution of unilateral feeding systems in snakes. Am Zool 41: Kochva E On the lateral jaw musculature of the Solenoglypha with remarks on some other snakes. J Morphol 110: Kochva E Development of the venom gland and trigeminal muscles in Vipera palestinae. Acta Anat 52: Koentges G, Lumsden A Rhombencephalic neural crest segmentation is preserved throughout craniofacial ontogeny. Development 127: Lakjer T Studien uber die Trigeminus-versorgte Kaumuskulatur der Sauropsiden. Copenhagen: C.A Reitzel. Langebartel DA The hyoid and its associated musculature in snakes. Ill Biol Monogr 38: Lee MSY Hidden support from unpromising data sets strongly unites snakes with anguiomorph lizards. J Evol Biol 22: Liem KF, Marx H, Rabb GB The viperid snake Azemiops: its comparative cephalic anatomy and phylogenetic position in relation to Viperinae and Crotalinae. Fieldiana Zool 59: Losos J, Hillis D, Greene H Who speaks with a forked tongue? Science 338: Luther A Uber die vom N. trigeminus versorgte Muskulatur der Amphibien, mit einen vergleichenden Ausblick uber den Adductor mandibulae der Gnathostomen, und einem Beitrag zum Verst andnis der organisation der Anuranlarven. Acta Societatis Scientiarum Fennicae 44: Matsuoka T, Ahlberg PE, Kessaris N, Ianarelli P, Dennehy U, Richardson WD, McMahon AP, Koentges G Neural crest origins of the neck and shoulder. Nature 436: McDowell S The architecture of the corner of the mouth in colubroid snakes. J Herpetol 20: McDowell S The skull of serpentes. In: Gans C, Gaunt AS, Adler K, editors. Biology of the reptilia, Volume 21. Ithaca, New York: Society for the Study of Amphibians and Reptiles. p Metscher B MicroCT for comparative morphology: simple staining methods alow high-contrast 3D imaging of diverse nonmineralized animal tissues. BMC Physiol. 9:11. Oelrich T The anatomy of the head of Ctenosaura pectinata (Iguanidae). Misc Publ Mus Zool Univ Michigan 94: Palot MJ, Radhakrishnan C First record of Yellow-bellied Sea Snake Pelamis platurus (Linnaeus, 1766) (Reptilia: Hydrophiidae) from a riverine tract in northern Kerala, India. J Threat Taxa 2: Rieppel O The external jaw adductor of amphisbaenids (Reptilia: Amphisbaenia). Rev Suisse Zool 86: Rieppel O The trigeminal jaw adductors of primitive snakes and their homologies with the lacertilian jaw adductors. J Zool 190: Rieppel O The development of the trigeminal jaw adductor musculature in the grass snake Natrix natrix. J Zool 216: Romer AS Osteology of the reptiles. Chicago: University of Chicago Press. Scanlon JD, Lee MSY The major clades of living snakes: morphological evolution, molecular phylogeny and divergence dates. In: Aldridge RD, Sever DM, editors. Reproductive biology and phylogeny of snakes. Enfield, New Hampshire: Science Publishers Inc. p Underwood G A contribution to the classification of snakes. London: Trustees of the British museum (Natural History). Wiens J, Hutter C, Mulcahy D, Noonan B, Townsend T, Sites JJ, Reeder T Resolving the phylogeny of lizards and snakes (Squamata) with extensive sampling of genes and species. Biol Lett 8: Zaher H Comments on the evolution of the jaw adductor musculature in snakes. Zool J Linn Soc 111: