cranial muscles of the anurans leiopelma hochstetteri and ascaphus truei and the homologies of the...

Cranial Muscles of the Anurans Leiopelma hochstetteriand Ascaphus truei and the Homologies of theMandibular Adductors in Lissamphibia and OtherGnathostomes

Peter Johnston*

Department of Anatomy with Radiology, University of Auckland, Auckland, New Zealand

ABSTRACT The frogs Ascaphus truei and Leiopelmahochstetteri are members of the most basal lineages ofextant anurans. Their cranial muscles have not beenpreviously described in full and are investigated here bydissection. Comparison of these taxa is used to review acontroversy regarding the homologies of the jaw adduc-tor muscles in Lissamphibia, to place these homologiesin a wider gnathostome context, and to define featuresthat may be useful for cladistic analysis of Anura. Anew muscle is defined in Ascaphus and is designated m.levator anguli oris. The differences noted between Asca-phus and Leiopelma are in the penetration of the jawadductor muscles by the mandibular nerve (V3). In thetraditional view of this anatomy, the paths of the trigem-inal nerve branches define homologous muscles. Thisscheme results in major differences among frogs, sala-manders, and caecilians. The alternative view is thatthe topology of origins, insertions, and fiber directionsare defining features, and the nerves penetrate the mus-cle mass in a variable way. The results given here sup-port the latter view. A new model is proposed for Lis-samphibia, whereby the adductor posterior (levatorarticularis) is a separate entity, and the rest of theadductor mass is configured around it as a folded sheet.This hypothesis is examined in other gnathostomes,including coelacanth and lungfish, and a possiblesequence for the evolution of the jaw muscles is demon-strated. In this system, the main jaw adductor in teleostfish is not considered homologous with that of tetrapods.This hypothesis is consistent with available data on thedomain of expression of the homeobox gene engrailed 2,which has previously not been considered indicative ofhomology. Terminology is discussed, and ‘‘adductor man-dibulae’’ is preferred to ‘‘levator mandibulae’’ to alignwith usage in other gnathostomes. J. Morphol.272:1492–1512, 2011. � 2011 Wiley Periodicals, Inc.

KEY WORDS: Lissamphibia; gnathostomes; cranialmusculature; homology; Ascaphus; Leiopelma

INTRODUCTION

The cranial musculature of frogs has been stud-ied in detail in many representatives of the majorclades, but incomplete data only are recorded onthat of the New Zealand frogs Leiopelma, whichare unique to this country. The descriptions of the

jaw muscles by Stephenson (1951) and Wagner(1934) are difficult to follow because of terminol-ogy, and are provided with limited illustrations.The muscles of the tongue and floor of the mouthhave, however, been examined (Trewavas, 1933).The jaw adductor muscles of the North Americanfrog Ascaphus have been described by Carroll andHolmes (1980) and Haas (2001). Horton (1982)used anuran tongue muscles for phylogeneticinference, including both these taxa. Ascaphus andLeiopelma are understood to be members of themost basal lineages of extant frogs (Gao andWang, 2001), but their relationships to each otherand to the rest of Anura have been debated. Mor-phologically based work has suggested that Asca-phus and Leiopelma are successive sister-taxa toall other frogs (Ford and Cannatella, 1993) on thebasis of a number of characters that unite Leio-pelma and all other frogs except Ascaphus. Gaoand Wang (2001), using osteological characters,placed Ascaphus and Leiopelma in the same clade.Pugener et al. (2003) did not include Leiopelma intheir larval morphology study and placed pipoidsand then Ascaphus as successive sister groups toall other frogs. Recent phylogenies based on orincluding molecular data (Frost et al., 2006; Bos-suyt and Roelants, 2009; Irisarri et al., 2010;Wiens, 2011) also place these frogs in the sameclade. Relatively few morphological synapomor-phies for an Ascaphus–Leiopelma grouping havebeen identified, and a number of features commonto these genera are retained plesiomorphies(Stephenson, 1951; Estes and Reig, 1973; Lynch,1973; Irisarri et al., 2010). Irisarri et al. (2010)

*Correspondence to: Peter Johnston, Department of Anatomywith Radiology, University of Auckland, Private Bag 92019, Auck-land, New Zealand. E-mail: [email protected]

Received 9 December 2010; Revised 8 April 2011Accepted 7 May 2011

Published online 15 August 2011 inWiley Online Library (wileyonlinelibrary.com)DOI: 10.1002/jmor.10998

JOURNAL OF MORPHOLOGY 272:1492–1512 (2011)

� 2011 WILEY PERIODICALS, INC.

suggested that the loss of the columella is the onlysynapomorphy of an Ascaphus–Leiopelma clade;however, Gao and Wang (2001) listed 10 furtherosteological features. The phylogeny of Ford andCannatella (1993) is based, for these taxa, on Can-natella (1985); this author points out thatAscaphus and Leiopelma are used as outgroups toall other frogs in his analysis, and that this makesidentification of autapomorphies and synapomor-phies for these taxa difficult. The absence of anonanuran outgroup also means that thephylogeny recovered, with Ascaphus and Leio-pelma as successive sister groups to other frogs, isthe only possible result. A full analysis of the mor-phological implications of an Ascaphus–Leiopelmaclade will not be attempted here, but would be of

considerable interest; features in the anatomystudied here will be summarized in a form suitablefor future cladistic use.

The homology of the different components of thejaw adductors among the Lissamphibian groupshas been a source of debate. ‘‘Homology’’ is usedhere in the sense of ancestral identity retainedduring evolutionary change, and ‘‘Lissamphibia’’refers to extant amphibians and their most recentcommon ancestor. The traditional interpretation ofthese muscles advanced by Luther (1914) definesmuscle sections as separated by the branches ofthe trigeminal nerve, such that the adductor (leva-tor) internus lies between the ophthalmic (V1) andthe maxillary (V2) divisions, the adductor externusbetween V2 and the mandibular nerve (V3), andthe adductor posterior is caudal to V3 (Fig. 1). Thisnomenclature of these muscles has been generallyaccepted from Luther’s study. Lubosch (1914)found that Luther’s nerve-based homology dis-torted an otherwise similar overall topology ofthese muscles in frogs and salamanders, but, asnoted by Haas (2001), Lubosch (1938) followed theLuther system without comment in his later work.Save-Soderbergh (1945) also followed the Lutherscheme, and his interpretation was developed fur-ther by Carroll and Holmes (1980) with the hy-pothesis that the opening of the skull table wasassociated with different muscles in the ancestorsof frogs and salamanders, with resulting differentmorphologies of the dorsal cranium. However,these authors did not specify what exactly theseproposed differences in skull opening were, andgiven the wide range of possible morphologies ofthe dorsal skull anatomy in both groups (Trueb,1993), it is not easy to understand how these dif-ferences in skull opening could be identified. Ior-dansky (1992) considered the topology of the jawadductor muscles in Lissamphibians, and proposedthat similarities of origin, insertion, tendon struc-ture, and relationship to other muscles are the fea-tures that define homologous muscles, with posi-tion of the nerve trunks not being an acceptablecriterion of homology. An alternative plan was putforward by Haas (2001), who looked at the devel-opment of the adductor muscles in anurantadpoles and compared it with examples of the sal-amander and caecilian clades. The variable path-way of V3 in closely related frogs led Haas to rejecta nerve-based homology, and he suggested arevised nomenclature based on topology (Fig. 1and Table 1).

The terminology applied to the mandibularadductors is somewhat problematic in that varioussystems have been applied and reflect differenthomology assumptions. Table 1 summarizes theterms used in important contributions to anuranjaw muscle literature in comparison with the sys-tem used here. The major difference among thesesystems is the designation of the muscle that

Fig. 1. Schematic dorsal views of jaw adductors in a typicalsalamander and a typical frog, with topographically similar oridentical muscles color coded and the maxillary (V2) and man-dibular (V3) nerves labeled. The nomenclature and homologyassumptions applied by various authors to the ‘‘longus’’ (orange)component of the jaw adductor are identified.

CRANIAL MUSCLES OF Leiopelma AND Ascaphus 1493

Journal of Morphology

arises from the dorsal, caudal margin of the skull,dorsal to the otic capsule. Luther (1914) inter-preted this component as adductor (levator) poste-rior in Anura, because it lies caudal to the mandib-ular nerve and as the adductor (levator) internusin Urodela, in which it lies rostro-medial to thisnerve. Haas (2001) interpreted this muscle sectionas the homologous ‘‘longus’’ in both taxa, takingthis term from Lubosch (1914).

An additional issue in nomenclature is the useof ‘‘levator’’ in amphibian literature (Edgeworth,1935; Duellman and Trueb, 1986; Haas, 2001),whereas ‘‘adductor’’ is typically used in publica-tions on other gnathostomes and by some authorson amphibians (Carroll and Holmes, 1980; Iordan-sky, 1992). The use of the term ‘‘levator’’ for themain elevator of the lower jaw appears to havebecome current following Edgeworth’s (1935) work.Edgeworth differentiated ‘‘levator’’ and ‘‘adductor’’in this way: when there is a holostylic upper jawsuspension (fusion of the palatoquadrate or itsremnant to the braincase) and (typically) absenceof a constrictor dorsalis muscle group (holocepha-lans, dipnoans, amphibians, and mammals) theterm ‘‘levator’’ is used for the elevator of the lowerjaw, but when a potentially mobile jaw suspensionand constrictor dorsalis muscles are present‘‘adductor’’ is used. This dichotomy was based onthe concept that the holostylic dipnoan–amphibiansituation is plesiomorphic, a view which was some-what unconventional at the time and has not beensupported by more recent knowledge (Groganet al., 1999). Another differentiation of ‘‘levator’’and ‘‘adductor’’ is that used by Vetter (1874) forfish in general and followed for basal teleost line-ages by Allis (1897, 1922) and Lauder (1980), inwhich ‘‘levator’’ is used for separate rostral compo-nents which may insert on the maxilla and‘‘adductor’’ for the remainder of the muscle mass.Yet another usage of ‘‘levator’’ and ‘‘adductor’’ isoffered by Rosen and Paterson (1969).

An alternative system of nomenclature for gna-thostome jaw adductors has been proposed byDiogo et al. (2008), in which components in tetra-pods are identified with a modification of the A1,A2, A3 system introduced for teleost fish by Vetter

(1874) and developed by Winterbottom (1974). Thishas not been used here, as I believe further com-parisons of the anatomy in all fish groups are nec-essary to determine if the teleost morphology isthe most appropriate model for a plesiomorphicstate, from which homologies in Lissamphibia andother tetrapods can be understood. In this study, Ihave examined taxa from other gnathostomegroups, to provide a phylogenetic framework inwhich the conclusions offered for Lissamphibia canbe placed.

The muscles of the floor of the mouth, tongue,and larynx in Anura have been documented formany species from various groups of frogs by Tre-wavas (1933), who included ‘‘Leiopelma hochstet-teri’’ but not Ascaphus truei. In fact, the New Zea-land frog examined by Trewavas was Leiopelmaarcheyi: this was noticed by Stephenson (1951),who recognized the difference in the branchialplate, and examined Trewavas’ specimen to con-firm it was L. archeyi. Subsequent citations of Tre-wavas (1933) and copies of her figures have not,however, recognized this error. L. hochstetteri hasa small alary process of the hyobranchial plate,which is absent in L. archeyi; a number of othermorphological differences among Leiopelma speciesare noted by Stephenson (1951) but none of theseappear to be relevant to this study. Horton (1982)included both genera in a comparison of tonguemuscles but restricted her report to two muscles,m. genioglossus and m. hyoglossus; the configura-tions of these muscles were used for phylogeneticinference. A brief preliminary description of the la-ryngeal muscles in L. hochstetteri and A. truei willbe given, and a full account will await a moredetailed investigation with three-dimensionalimage reconstruction and a consideration of laryn-geal function. Terminology used for the tongue,floor of mouth, and laryngeal muscles follows thatof Trewavas (1933), with the exception that ‘‘m.constrictor laryngis’’ is used here to refer to the la-ryngeal constrictor, which encircles the larynxwithout origin from the branchial cartilage.

The m. levator bulbi, as part of the trigeminalmuscle group, has been included, but the externalocular muscles have not been examined; there is

TABLE 1. Synonyms for jaw adductors in Anura

This work Luther (1914)Edgeworth

(1935)Carroll and

Holmes (1980)Iordansky(1996) Haas (2001)

A.m. internus(rostralis, caudalis)

Pterygoideus L.m. anterior A.m. internus Pseudotemporalis L.m. internus

A.m. longus A.m. posterior longus L.m. posterior A.m. posterior longus A.m. temporalis L.m. longusA.m. externus A.m. externus 1 a.m.

posterior subexternusL.m externus A.m. externus, a.m.

posterior subexternusA.m. externus L.m. externus

A.m. lateralis A.m. posterior lateralis L.m. anteriorlateralis

A.m. posterior lateralis A.m. lateralis L.m. lateralis

A.m. posterior A.m. posteriorarticularis

L.m. anteriorarticularis

A.m. posterior articularis A.m. posterior L.m. articularis

A.m., adductor mandibulae; L.m., levator mandibulae.

1494 P. JOHNSTON


very little comparative literature on these musclesin Lissamphibia.

The aims of this work are to describe the mor-phology of the cranial muscles in these taxa, to listimportant differences as possible characters forfuture phylogenetic analysis, to use the compari-son of these taxa to investigate the homologies ofthe jaw muscles, and to place the resulting homol-ogy proposal in a phylogenetic context.

METHODS

Anuran material: L. hochstetteri: four specimens were used,Otago University, Department of Anatomy collection L06, L08,L09; Auckland Museum AIM LH16. A. truei: two specimens,Auckland Museum AIM LH82 and AIM LH83.Comparative material (numbers in brackets): holocephalan:

Callorhinchus milli (elephant shark), National Museum of NewZealand NMNZ P29678; elasmobranch: Cephaloscyllium isabel-lum (New Zealand carpet shark, a scyliorhinid or cat sharkrather than a true carpet shark; 2) commercial supplier; teleost:Oncorhynchus tschawtscha (Chinook salmon; 2), commercialsupplier; coelacanth: Latimeria chalumnae, magnetic resonanceimaging (MRI) scan images supplied by Digital Fish Library,University of California, San Diego (http://digitalfishlibrary.org),specimen: Scripps Institute of Oceanography SIO 75-347, andcomputed tomography (CT) scan images supplied by the Univer-sity of Texas Digital Morphology Group (http://digimorph.org),specimen: American Museum of Natural History AMNH 32949;lungfish: Neoceratodus forsteri (Australian lungfish; 3) Mac-quarie University, NSW, Australia, unnumbered, and MRI scan,Digital Fish Library, specimen: California Academy of SciencesSU 18139; rhynchocephalian: Sphenodon punctatus (tuatara; 4),Otago University Department of Anatomy collection: tuatara A,B, C, D; turtle: Lepidochelys olicacea (Olive Ridley sea turtle),Auckland Museum AIM Tax 09-012; crocodilian: Alligator mis-sissippiensis (American alligator; 2) Macquarie University,unnumbered; mammal (marsupial): Trichosurus vulpecula(brush-tailed possum; 2), personal collection, unnumbered. Pos-session of specimens of protected species was in accordancewith New Zealand legislation.Specimens were dissected under dissecting microscope magni-

fication on both sides of the head. The findings were drawnfreehand and photographed; for A. truei and L. hochstetteri,results were transcribed onto skull outlines drawn from CTscan images provided by the University of Texas Digital Mor-phology Group (http://www.digimorph.org) using Corel GraphicsSuite X4 (specimens: A. truei: University of Michigan Museumof Zoology UMMZ 133050; L. hochstetteri: University of Michi-gan Museum of Zoology UMMZ 146857). The larynx was dis-sected from both ventral and oral approaches. Computer modelsof the jaw adductor muscles in Lissamphibians were made withMaya 2009 (Autodesk) by drawing schematic three-dimensionalshapes representing the muscle topologies according to the hy-pothesis presented below. For brevity, sections of the m. adduc-tor mandibulae will be referred to as internus, longus, externus,and lateralis (see Table 1). Terminology for the muscles of theoccipital end of the anuran skull is from Ritland (1955), for cra-nial muscles in holocephalans from Didier (1995), and forsharks from Mallatt (1997). Criteria used to determine homol-ogy of the cranial muscles, in decreasing order of importance,were: evolutionary continuity of morphological topology on anacceptable cladogram, skeletal connections, innervation, geneexpression, and morphological ontogeny.

RESULTSM. Adductor (Levator) Mandibulae

L. hochstetteri (Fig. 2A–C): this is a continuousblock of muscle with no distinct separation into

individual muscles apart from a plane between them. adductor mandibulae posterior and the rest ofthe muscle group, and a separate m. adductormandibulae lateralis. The insertion of the adductorgroup on the mandible is depicted as the internus,longus, externus, and lateralis muscles wrappedaround the insertion of m. adductor posterior in aU shape (Fig. 3). As in other frogs, the most ros-tral origin is from the dorsal braincase medial tothe eye. This origin extends far rostrally, furtherthan in A. truei and in most other frogs. The inter-nus muscle (see Table 1 for nomenclature) is com-posed of two parts merging into one another: a ros-tral part (Fig. 2B,C: ir), in which the fibers con-verge onto a flat tendon which inserts of themandible immediately medial to its articulationwith the quadrate, and a caudal part (Fig. 2B,C:ic) with a more ventral fiber direction and a fleshyinsertion onto the medial side of the mandible ad-jacent to the tendon of the rostral section. Caudalto the internus, the longus section of the musclearises from the parietal region of the braincase,from the convexity of the otic capsule as far thedorsal and caudal limits of the skull, including thecrista parotica. These fibers pass ventrally androstrally, medial to the processus zygomaticus ofthe squamosal, to insert around the apex of the Uon the coronoid area of the mandible (Fig. 3). Con-tiguous laterally with this muscle is the externus,arising from the squamosal and quadrate andinserting as the lateral limb of the U. The lateralisis a small separate body of muscle at the most lat-eral and caudal extent of the group, arising fromthe quadrate just dorsal to its articulation withthe mandible and passed ventral and rostral to lielateral to the V3. The adductor posterior is a cylin-drical muscle body, passing from the anterior faceof the quadrate to the dorsal aspect of the mandi-ble immediately rostral to the articulation.

The course of V2 and V3 can be seen in Figure2A–C: V2 appears between the internus and longusmuscles and follows the curve of the orbit aroundthe rostral border of the adductor muscle block toreach the maxilla. V3 passes through the adductorgroup lying on the rostral face of adductor poste-rior to emerge laterally at the caudal edge of theexternus component and is then covered for ashort distance by the lateralis.

A. truei (Fig. 2F,G): the general configuration ofthe adductor muscle group is similar to L. hoch-stetteri in that the m. adductor posterior (articu-laris) is separate, but the rest of the muscle is acontiguous block; the major differences are thatthe origin of the internus does not extent so farrostrally on the braincase, and that the mandibu-lar nerve is in a different situation: in A. truei(Fig. 2F,G), V3 emerges between the internus andlongus muscles together with V2, curls around therostral face of the longus and externus andthen, enters the muscle block to pass within the



externus and lateralis sections to emerge again atthe dorsal border of the mandible. A tendon isagain present in the internus muscle and forms

the only insertion of this section, similarly immedi-ately medial to the quadrate-mandibular articula-tion. The origin of the longus, dorsal to the otic

Fig. 2. (A) Leiopelma hochstetteri, cranial muscles in left lateral view, superficial exposure. (B) L. hochstetteri, deeper exposure,showing internus and lateralis muscles, and adductor posterior with maxillary and mandibular nerves. (C) L. hochstetteri, dorsalview, left side: superficial exposure, right side: deeper components of jaw adductor with maxillary and mandibular nerves. (D) L. hoch-stetteri, left dorsolateral view of skull with oto-occipital muscles, orientation for magnified square in (e). (E) L. hochstetteri, magnifiedleft oto-occipital region showing origins of mm. petrohyoidei and the medial bundle of m. depressor mandibulae. Gray shading for per-spective. (F) Ascaphus truei, cranial muscles in left lateral view, superficial exposure. (G) A. truei, dorsal view, left side: superficial ex-posure, right side: deeper components of jaw adductor with m. levator bulbi, maxillary and mandibular nerves. Specimen details: (A)–(E) muscles and nerves from AIM LH16, skull outline from UMMZ 146857; F–G, muscles and nerves from AIM LH82, skull outlinefrom UMMZ 133050. Abbreviations: c-o, origin of m. cucullaris; dm, m. depressor mandibulae; dm-m, medial part of m. depressormandibulae, arising in cranio-quadrate passage; dm-o, origin of m. depressor mandibulae; dmp, m. depressor mandibulae posterior;dn, m. depressor nictitantis; ds, m. dorsalis scapulae; e, m. adductor mandibulae externus; i, m. adductor mandibulae internus; ic, cau-dal section of internus; ir, rostral section of internus; lat, m. adductor mandibulae lateralis; lao, m. levator anguli oris; lb, m. levatorbulbi; lig, ligament connecting processus zygomaticus and lower eyelid, insertion of m. levator bulbi; lo, m. adductor mandibulae lon-gus; lsi, m. levator scapulae inferioris; lss, m. levator scapulae superioris; m, mandible; p, m. adductor mandibulae posterior; r, rictalplate; rha, m. rhomboideus anterior; rhp, m. rhomboideus posterior; ss, suprascapula; V2, maxillary nerve; V3, mandibular nerve.

1496 P. JOHNSTON


capsule, is similar in these two frogs. A new mus-cle is identified in A. truei and is here named m.levator anguli oris (Fig. 2F,G: lao). This small mus-cle arises from the dorsal surface of the palatalmembrane in the triangular area between ptery-goid and maxilla, and wraps around the rostralborder of the longus muscle, to pass medial to themaxilla and maxillo-quadrate ligament and inserton the rictal plate at the corner of the mouth. Thisinsertion is typical of the m. levator anguli oriscommon in Lepidosauria, for example, Sphenodon(Haas, 1973), and this new muscle in A. truei isnamed accordingly.

M. Levator Bulbi

Findings were similar in both frogs and areillustrated for A. truei (Fig. 2G). A thin sheet ofmuscle arises from the interorbital septum andcaudal aspect of the nasal capsule and passes ven-tral to the eye within the orbital fascia, and

inserts on a ligamentous thickening of the junctionof the orbital fascia with the thick fascia coveringthe mandibular adductors. This ligament, which ismore pronounced in L. hochstetteri than in A.truei, passes from the tip of the processus zygoma-ticus of the squamosal ventrally toward the caudalend of the maxilla, where it fades out into the fas-cial structure of the lower lid (see Fig. 2A,F). Aslip of muscle separates from the m. levator bulbias the m. depressor membranae nictitantis ofLuther (1914), inserting into the lower lid and nic-titating membrane.

M. Depressor Mandibulae

L. hochstetteri (Fig. 2A–C): two components ofthis muscle are present and are designated m.depressor mandibulae and m. depressor mandibu-lae posterior after Kleinteich and Haas (2011). Therostral portion originates from the quadrate, fromthe depths of the cranio-quadrate fossa (Fig. 2E),from the fascia covering the m. adductor mandibu-lae longus, and from the crista parotica along theposterior margin of the skull. The m. depressormandibulae posterior arises from the fascia of them. dorsalis scapulae. These muscles insert on theretroarticular process of the mandible with thecaudal part lateral to the rostral.

A. truei (Fig. 2F,G): the depressor muscles aresimilar to those of L. hochstetteri, except that anorigin within the cranio-quadrate fossa was notseen, and the insertion of the components onto themandible is reversed, that is, the rostral part islateral.

Muscles in the Floor of the Mouth andTongue

Terminology of the various features of the bran-chial plate is indicated in Figure 4A,B, which alsoshows sites of muscle attachment.

L. hochstetteri (Fig. 5A,B): the findings are muchas described by Trewavas (1933). The most ventrallayer comprises mm. intermandibularis posterior,intermandibularis anterior (mentalis), and inter-hyoideus, all of which are typical for frogs. Mm.intermandibularis posterior and interhyoideusform a continuous sheet at the midline and areseparate at their origins from the mandible andthe dorsal extension of the hyoid cartilage, respec-tively. The m. intermandibularis anterior is asmall but thicker body of fibers lying superficial tom. intermandibularis posterior. Deep to the inter-mandibularis–interhyoideus layer, in ventral view,are seen the mm. geniohyoideus and sternohyoi-deus converging on the central shield-shaped partof the branchial cartilage plate. The m. geniohyoi-deus takes the form of two adjacent muscle belliesseparated by the hypoglossal nerve, inserting onthe central part of the branchial plate near the

Fig. 3. Schematic cross-section of left jaw adductors in dor-sal view, showing components of the muscle as a sheet foldedaround m. adductor posterior, with courses of nerves indicated.The gray bands connecting the muscles indicate gaps betweenthem, emphasizing the continuous S-shaped cross-section.Abbreviations: V2, maxillary nerve; V3, mandibular nerve.



rudimentary alary process; together these musclebellies represent only the lateral component of thetwo-part m. geniohyoideus found in other frogs.The m. sternohyoideus arises from the dorsal sur-face of the caudal end of the sternum and insertsin the branchial plate adjacent to the m. genio-hyoideus. Medial to these muscles is the m. hyo-glossus, originating from the ventral face of the tipof the posteromedial process of the branchial plate.Lateral to these longitudinally oriented muscles(mm. geniohyoideus, sternohyoideus and hyoglos-sus), cranial nerve IX, the m. omohyoideus, andthe most rostral of the four mm. petrohyoidei (mm.levatores arcuum branchialium of Ziermann andOlsson, 2007) are present. Dorsal to m. geniohyoi-deus are the interdigitations of m. hyoglossus andm. genioglossus; the latter muscle arises from therostral end of the mandibular arc, with cranialnerve XII ramifying among the digitations. In the

specimens examined here, four digitations of eachmuscle are seen in two specimens, three in another,and four including one very small digitation inanother. These digitations are stacked on eachother in a series of latero-medial planes (Fig. 5B).On removal of the m. hyoglossus, the ventral aspectof the m. constrictor laryngis is present immedi-ately caudal to the posteromedial processes of thecartilage. The mm. petrohyoidei I–IV arise from thecaudal aspect of the prootic (Fig. 2E) and I–IIIinsert on the branchial plate as shown in Figure4A; the mm. petrohyoidei IV converge on a raphein the midline close to the convergence of the bron-chial processes of the cricoid (Fig. 4A).

A. truei: these muscles (Fig. 5C) are generallysimilar to those of L. hochstetteri. The m. inter-mandibularis anterior is a smaller body of muscleand lies on the symphysial region of the mandiblesrather than partly overlying m. intermandibularis

Fig. 4. Ventral view of branchial plate with muscle attachments and larynx with muscles: (A) Leiopelma hochstetteri. Musclesfrom AIM LH16, branchial plate outline after Stephenson (1951). (B) Ascaphus truei. Muscles from AIM LH82, branchial plate out-line after Frazier (1924). Dorsal view of caudal end of branchial plate and larynx with muscles: (C) L. hochstetteri. AIM LH16. (D)A. truei. AIM LH82. Shading: medium gray, branchial cartilage; dark gray, cricoid cartilage; light gray, bone. Abbreviations: a, ary-tenoid cartilage; al, alary process (absent in Ascaphus); c, cricoid; cd, m. constrictor laryngis dorsalis; cl, m. constrictor laryngis; cv,m. constrictor laryngis ventralis; dl, m. dilator laryngis; gh, m. geniohyoideus—insertion; hc, hyoid cornu; hg, m. hyoglossus—ori-gin; oh, m. omohyoideus—insertion; ph, parahyoid ossification; pl, postero-lateral process of branchial plate; pm, postero-medialprocess of branchial plate; p1–4, mm. petrohyoidei 1–4 (insertions in B); sh, m. sternohyoideus—insertion.

1498 P. JOHNSTON


posterior as in L. hochstetteri. The m. interhyoideusis separate from m. intermandibularis posterior. Inthe deeper layers, m. geniohyoideus is a single, lat-

eral muscle belly, covering the hypoglossal nerve,and inserts on the lateral edge of the hyoid plate asin L. hochstetteri, the alary process being absent inA. truei. The m. omohyoideus is absent, and themm. petrohyoidei are positioned to insert more ros-trally than in L. hochstetteri. The origins of themm. petrohyoidei are similar in A. truei and L.hochstetteri, and their insertions on the branchialplate are shown in Figure 4B. Three interdigita-tions of the mm. hyoglossus and genioglossus arepresent and are stacked as a series of dorso-ventralplanes (Fig. 5C). Between the posteromedial proc-esses of the hyoid apparatus, the insertion of m.petrohyoideus IV meets and fuses with the ventralpart of m. constrictor laryngis (Fig. 4D).

Muscles and General Morphology of theLarynx

L. hochstetteri (Fig. 4A,C): the larynx is typical ofthe anuran pattern (Hermida and Farias, 2009)with a smaller dorsal chamber adjacent to the aryte-noid cartilages, separated from the larger ventralchamber by the vocal chords. The laryngeal musclescomprise the dilator and constrictor: m. dilator lar-yngis originates from the postero-medial process ofthe branchial plate and inserts on the arytenoid car-tilage rostrally. The constrictor laryngis complexcomprises several muscles: a cylindrical muscle (m.constrictor laryngis), which surrounds the larynxand is interrupted by dorsal and ventral raphes, andseparate dorsal and ventral muscles (mm. constric-tor laryngis dorsalis and ventralis), which arisefrom the postero-medial process and reach the dor-sal and ventral raphes, respectively.

A. truei (Fig. 4B,D): the morphology of the lar-ynx differs from L. hochstetteri in that the aryte-noids and dorsal chamber of the larynx are verysmall. Vocal chords are present as narrow ridges(0.7 mm) projecting medially from the ventral edgeof the arytenoids. The arrangement of muscles isof a simple form: m. dilator laryngis arises fromthe postero-medial process of the branchial plateand inserts at the rostral end of the arytenoid

Fig. 5. Ventral views of floor of mouth and tongue muscles.(A) Leiopelma hochstetteri: left side (right side of picture), su-perficial exposure; right side, exposure after elevation of mm.intermandibularis and interhyoideus. Specimen: AIM LH82. (B)L. hochstetteri: right side (left side of picture) as in right side ofA above; left side, exposure after removal of m. geniohyoideusand a portion of the chest wall. AIM LH82. (C) Ascaphus truei:right side (left side of picture), exposure after elevation of mm.intermandibularis and interhyoideus; left side, exposure afterremoval of m. geniohyoideus. AIM LH16. Abbreviations: c, cri-coid cartilage; d, m. deltoideus; gg, m. genioglossus; gh, m. gen-iohyoideus; hc, hyoid cornu; hg, m. hyoglossus; ih, m. interhyoi-deus; ima, m. intermandibularis anterior; imp, m. intermandi-bularis posterior; m, mandible; oh, m. omohyoideus; p1–4, mm.petrohyoidei1–4; ps, m. pectoralis, sternal head; sh, m. sterno-hyoideus; IX, glossopharyngeal nerve; XII, hypoglossal nerve.



cartilage. The m. constrictor laryngis is present asa single muscle, which arises from a midline raphecaudal to the laryngeal aditus and surrounds thelarynx on each side to meet as a relatively narrowrostral raphe rostral to the aditus and as a contin-uous muscle ventral to the larynx.

Summary of Morphological Features

Significant differences and potential synapomor-phies of these two anurans are set out in Table 2,with a salamander (Siren lacertina) and an‘‘advanced’’ frog (Rana esculenta) for comparison,in a form that makes these data available for clad-istic analysis. Proposed plesiomorphic states areindicated.

Comparative Material: Jaw Adductors inOther Gnathostomes

Accounts of jaw adductor morphology for thetaxa examined, or closely related taxa, are avail-able in the literature (see references in Table 3);the findings of the this study are summarized on a

simplified cladogram in Figure 7. This shows dor-sal views of the insertions of the jaw adductors onthe left mandible, as if projected onto a two-dimen-sional plane. This enables comparison of Lissam-phibia and other gnathostomes in relation to thehypothesis presented below, in which part of thejaw adductor is seen as a folded sheet. Muscleinsertions are color coded according to homologyproposals discussed below, and Table 3 lists namesof jaw muscles in the various taxa according to thesame system. It must be noted that the ‘‘mandible’’of chondrichthyan taxa is not necessarily homolo-gous with that of osteichthyans, in which dermalbones have been applied to the mandibular(Meckel’s) cartilage.

Several findings are presented in more detail:the interpretations of the anatomy in lungfish andcoelacanth are different from that in recent publi-cations (Anderson, 2008; Diogo et al., 2008; Diogoand Abdala, 2011). Specific findings in the sharkare important for homology considerations, andthe shark also presents most of the features of theplesiomorphic gnathostome state to be discussedbelow.

TABLE 2. Comparison of findings with Siren lacertina (after Luther, 1914) and Rana esculenta (after Gaupp, 1896)

Sirenlacertina

Ascaphustruei

Leiopelmahochstetteri

Ranaesculenta

M. levator bulbi, multiple sections(excluding depressor nicitantis)

Yes No No Yes

vocal sacs Noa No No NoM. geniohyoideus, lateral and medial Mediala,b Lateral Lateral Medial and lateralM. depressor mandibulae origin, includes

medial face of quadrateNoa No Yes No

M. depressor mandibulae posterior insertion,medial to insertion of depressor mandibulae

Yesa,c No Yes Yes

Ligamentous caudal attachment of m. levator bulbi No Yes Yes NoMuscle attached to rictal plate—angle of mouth No Yesa,d No NoM. adductor mandibulae internus insertion

by long tendonYesa Yes Yes Yes

Discrete rostral component of m. adductormandibulae internus

Noa No Yes No

V2 lies rostral and medial to m. adductormandibulae longus

No Yesa,e Yes Yes

V3 lies rostral and medial to m. adductormandibulae longus

No Yes No Yes

M. a.m.f lateralis contiguous with m. a.m. externus — Yes No NoInterdigitations of mm. genioglossus and

hyoglossus— 3 3–4 2

M. omohyoideus, present — No Yes YesInsertion m. petrohyoideus IV on branchial plate Yesg Noh Noi Yesj

M. interhyoideus contiguous with m.intermandibularis posterior

Yes Yes No Yes

M. constrictor laryngis, one muscle belly Yesa,k Yes No No

aProposed plesiomorphic states.bJarvik (1963).cBauer (1997).dSee text, anguli oris muscles present in some salamanders and in the lungfish Protopterus and Lepidosiren (Edgeworth, 1935).eV2 lies in this position in lungfish, see Figure 6D.fM. a.m. 5 m. adductor mandibulae.gCeratobranchialia 3 1 4.hFuses with constrictor larynges.iMeets opposite muscle in midline caudal to larynx.jPosteromedial process 1 cricoid.kThe larynx is absent in plethodontid salamanders.

1500 P. JOHNSTON


C. isabellum (New Zealand carpet shark; Fig.6A,B): the mm. adductores arcuum branchialiumare similar in configuration to the m. adductormandibulae, in that muscle bellies originate fromthe epibranchial and ceratobranchial cartilages(palatoquadrate and mandibular cartilages in thecase of the mandibular arch) and converge on ahorizontal intermediate tendon, the Zwischen-sehne of Luther (1909). Thus, the m. adductormandibulae presents as a greatly enlarged serialhomolog of an m. adductor arcus branchialis. Them. preorbitalis is a separate rostral muscle, arisingfrom the ethmoid and nasal cartilages and insert-ing on the Zwischensehne. In general, the cranialmuscles are similar to the descriptions of smaller,unspecialized squaliform sharks by Luther (1909)and Kesteven (1942–1945). Features commonlyfound in carchariniforms (to which this taxonbelongs) according to Compagno (1988), such as aneyelid muscle and a divided origin of m. preorbita-lis, have not been found.

N. forsteri (Australian lungfish; Fig. 6D): thechondrocranium persists into the adult, and thereis little ossification of the skull. Nomenclature forcranial elements follows Bartsch (1994). The m.adductor mandibulae presents contiguous medialand lateral sections, which are more separate andaligned rostral and caudal, respectively, in thelarva (Ericsson et al., 2011) and are thus namedm. adductor mandibulae anterior and posterior.In the adult, these muscles arise from the fusedpalatoquadrate and prootic cartilages and fromthe extensive dorsal skeletal covering of theadductor fossa (the lateral fronto-parietal) almostto the caudal limit of the skull, where they areseparated from the insertions of the epaxial mus-cle group by a transverse ridge. V2 and V3 passbetween these two sections of the adductor. Them. adductor mandibulae anterior converges in afan-shaped fashion onto a stout tendon thatinserts on the Meckel’s cartilage and the os man-dibularis internus (a medial dermal bone) at the

TABLE 3. Homology of jaw adductors in gnathostomes



preglenoid eminence. The most medial and ven-tral fibers pass almost directly lateral from thepterygo-palatinum to this tendon. The m. adduc-tor mandibulae posterior has a fleshy insertionlateral on the lateral edge of Meckel’s cartilage,and also sends a tendon to fuse with that of m.adductor mandibulae anterior.

L. chalumnae (coelacanth; Fig. 6C): three sec-tions of the m. adductor mandibulae are seen onthe MRI scans and correspond with the dissectionfindings of Millot and Anthony (1958) and Forey(1998; pp. 198–199). The most medial part (chefantero-superior of Millot and Anthony, 1958), with

vertically oriented fibers, arises from the antoticprocess of the ethmosphenoid and also from thedorsal edge of the metapterygoid; this muscleinserts by a tendon onto the medial edge of Meck-el’s cartilage. A central portion (chef moyen) ariseson the dorso-caudal rim of the palatoquadrate car-tilage and pass ventrally and rostrally to join thefibers of the most lateral portion (chef postero-infe-rieur), which is a large muscle originating alongthe caudal edge of the palatoquadrate and fillingthe Meckelian fossa, flanked medially by the prin-cipal coronoid and laterally by the angular bone,to insert on Meckel’s cartilage.

Fig. 6. (A) Cephaloscyllium isabellum (scyliorhinid shark). Left lateral view of cranial musculature, superficial exposure, withthe covering of the second gill-chamber removed. (B) Magnified view of the square in (A) to show details of deeper muscles. (C) Lat-imeria chalumnae (coelacanth). Axial (coronal) MRI scan image with jaw adductors. The outline of the tendon of the antero-supe-rior adductor muscle is enhanced. Specimen: SIO 75-347. (D) Neoceratodus forsteri (Australian lungfish). Left lateral view of jawadductors after removal of their fronto-parietal skeletal covering. Abbreviations: aab, m. adductor arcus branchialis; am, m. adduc-tor mandibulae; am-as, m. adductor mandibulae, antero-superior part; am-pi, m. adductor mandibulae, postero-inferior part; am-pi-im, intramandibular part of am-pi; ama, m. adductor mandibulae anterior; amp, m. adductor mandibulae posterior; bc, branchialcartilage; ep, epaxial spinal muscles; fp, cut edge of fronto-parietal covering of adductor compartment; hy, m. hyomandibularis; ib,m. interbranchialis; ig, internal gill opening; gs, gill slit; lc, labial cartilage; lhy, m. levator hyomandibularis; lpq, m. levator palato-quadrati; m, mandible (Meckel’s cartilage); ms, m. spiracularis; po, m. preorbitalis; s, spiracle; sc, superficial constrictor muscle; V2,maxillary nerve, V3, mandibular nerve; zs, Zwischensehne.

1502 P. JOHNSTON


Fig. 7. Gnathostome cladogram, showing insertions of left jaw adductors on the mandible in dorsal view, color coded accordingto homologies with same colors as Figures 1 and 3. Polypterus bichir after Allis (1922), Ichthyophis glutinosus a caecilian, afterLuther (1914). Other taxa, see ‘‘Methods’’ section for details. Abbreviations: am, m. adductor mandibulae; am-as, m. adductor man-dibulae, antero-superior part; am-pi, m. adductor mandibulae, poster-inferior part; ama, m. adductor mandibulae anterior; ame, m.adductor mandibulae externus; ami, m. adductor mandibulae internus; amp, m. adductor mandibulae posterior; ao, mm. levatorand retractor anguli oris; lao, m. levator anguli oris; las, m. labialis superior; lms, m. levator maxillae superioris; po, m. preorbita-lis; ps, m. pseudotemporalis; pt-a, m. pterygoideus atypicus; pt-d, m. pterygoideus dorsalis; pl-l, m. pterygoideus lateralis; pt-m, m.pterygoideus medialis; pt-v, m. pterygoideus ventralis; tp, m. tensor palati; tt, m. tensor tympani.



DISCUSSIONMandibular Adductors and Depressors

The anatomy of the mandibular adductors in L.hochstetteri and A. truei resembles that of otherfrogs (Luther, 1914) in overall arrangement. Thedifferences to be noted are the rostral extent of theorigin of this muscle group in L. hochstetteri andthe location of the mandibular nerve as it passesthrough the muscle in both taxa. The situation inA. truei, in which the nerve passes across the ros-tral face of the longus section of the muscle (Fig.2F,G), is similar to that found in most frogs (Haas,2001). The passage of V3 caudal to the longus sec-tion as seen in L. hochstetteri is seen also in themicrohylids Gastrophryne carolinensis and Ela-chistocleis bicolor; however, in other microhylids(Kaloula pulchra and Scaphiophryne madagascar-iensis), this nerve lies in the more usual rostralposition (Haas, 2001). This variation amongrelated taxa with otherwise similar jaw adductorsindicates that the nerve course is not necessarilythe defining feature in homology, and led Haas(2001) to reject the Luther (1914) paradigm.

Carroll (2007, 2009) attempted to dismiss Haas’proposal, based on a perceived difficulty in identify-ing small nerves in these larval forms, but this crit-icism becomes less relevant when two importantlines of evidence are noted. First, the finding ofHaas (2001) that the relations of muscle bodies andtrigeminal branches are retained through metamor-phosis into adult frogs was also noted by Schlosserand Roth (1995), and second, various courses of themandibular nerve through the adductor musclegroup in adults had been previously documented byStarrett (1968), with results identical to the larvaland adult results of Haas. Starrett (1968) alsoreported a number of microhylid frogs in which V2

passes through rather than rostral to the longussection of the muscle, similar to the situation in sal-amanders; this observation also strengthens theargument that placement of nerve trunks is not thedefining feature of muscle identity in Lissamphi-bians. Cannatella (1999) also chose to disregard tri-geminal nerve location in defining the mandibularmuscle components in anuran larvae.

The findings reported here support the positionof Iordansky (1992, 1996) and Haas (2001): homol-ogy among parts of the anuran m. adductor mandi-bulae is better defined by topological featuresother than relations to branches of the trigeminalnerve. The differences noted in the course of themandibular nerve between L. hochstetteri and A.truei and across Lissamphibia in general are out-weighed by other similarities in muscle topology.In particular, the ‘‘footprint’’ of insertion of theadductor group on the mandible is similar amonganurans and caudates, if nerve-based criteria aredisregarded, and this pattern of insertion can alsobe recognized in caecilians (Fig. 7).

Further examples of cranial morphology withtrigeminal nerve courses contravening the Lutherprinciple can be given. Haas (2001) cited the dataof Friel and Wainwright (1997) on tetraodontiformfishes with trigeminal nerve course differencesamong related taxa, and a similar variability isnoted in galaxioid fishes by Williams (1997). Edge-worth (1935) and Winterbottom (1974) bothemphasized that the course of V3 is not a reliablemarker for division of components of the jawadductor complex in teleosts, although Gosline’s(1989) interpretation of teleost muscle homologiesgives the nerve course a somewhat more reliablestatus. The complex relation of the mandibularnerve and the mandibular adductors in the basalteleost fish Polypterus bichir (Allis, 1922) does notconform to Luther’s plan. However, these compari-sons assume that the adductor muscles in teleostsand tetrapods are homologous, and this will bequestioned below. In tetrapods, Holliday andWitmer (2007) described the course of the maxil-lary nerve in crocodilians and conclude that the m.pseudotemporalis is defined by other features andlies medial to this nerve, unlike the lateral situa-tion in other sauropsids; within Lissamphibia, thecourse of the V1 trigeminal division is lateral tothe internus component of the adductor in S. lacer-tina, instead of the medial path in all other taxadescribed (Luther, 1914). Carroll and Holmes(1980) considered that the adductor pattern andthe inferred differences in fenestration of the der-mal skull table in frogs and salamanders are suffi-ciently divergent as to preclude their origin fromthe same temnospondyl lineage, but in more recentwork, Carroll (2009) accepts the origin of bothgroups from dissorophoid temnospondyls.

Another result of Haas’ (2001) proposal refers tolarval forms of some frogs in which a portion ofthe externus component is separated off by themandibular nerve and traditionally referred to as‘‘subexternus’’; Haas proposes that this distinctionis abandoned in favor of a single externus musclewhich is penetrated by V3 in some taxa; this seemslogical and is in keeping with the homology pro-posal offered here.

Hypothesis—Organization of LissamphibianMandibular Adductors

Here, the author offers a new hypothesis, whichextends the homology conclusion of Haas (2001):the m. adductor mandibulae posterior (articularis)is a separate and constant entity, and the remain-der of the adductor block can be visualized as afolded sheet, which wraps around the adductorposterior, and is variably separated into discretemuscles. The branches of the trigeminal nerve tra-verse the muscle group but do not define its com-ponents. This hypothesis is advanced for anuranswith the data presented here and from Luther

1504 P. JOHNSTON


(1914), Starrett (1968), and Haas (2001), for sala-manders from Luther (1914), Carroll and Holmes(1980), Haas (2001), and Lubosch (1938), and incaecilians from Luther (1914) and Kleinteich andHaas (2007). The formulation of this plan for cae-cilians accepts Haas’ (2001) interpretation that thecaecilian m. levator mandibulae ‘‘externus’’ ofEdgeworth (1935) and ‘‘medius’’ of Luther (1914)represent the posterior (articularis) component asseen in the other Lissamphibia. The identity of theposterior adductor as a structure independent ofthe rest of the adductor is deduced from its con-stant presence in Lissamphibia and in tetrapodsapart from mammals (Table 3 and Holliday, 2009),

passing between the palatoquadrate and Meckel’scartilage, or their ossifications, immediately rostralto their articulation. In mammals, this muscle hasbecome the mm. tensor tympani and tensor palati(Edgeworth, 1935). The mandibular nerve passesrostral to the posterior adductor in all tetrapods.The configuration of the folded sheet in adult Lis-samphibia is portrayed in Figures 3 and 8. InAnura and Caudata, this morphology takes theform of a double fold with an S-shaped horizontalsection, and in Gymnophiona a single fold and aU-shaped section.

A folded or twisted sheet morphology can also berecognized in many of the figures of Luther (1914)

Fig. 8. Left lateral views of 3D models illustrating the m. adductor mandibulae posterior as a separate muscle, and the rest ofthe adductor placed around it as a folded sheet, with nerve trunks in situ and the caecilian m. pterygoideus depicted as a separatedportion on the sheet in Ichthyophis glutinosus (after Iordansky, 1996). The salamander Salamandra salamandra is after Iordansky(1996). Abbreviations: e, externus; i, internus; l, longus; lat, lateralis; p, m. adductor mandibulae posterior; pt, m. pterygoideus; V2,maxillary nerve; V3, mandibular nerve.



in which the isolated adductors and mandible areportrayed from a medial view. In caecilians with aseparate ‘‘pterygoideus’’ section of the internusmuscle such as Ichthyophis glutinosus and Siphon-ops annulatus (Luther, 1914), this can be inter-preted as a section of the sheet missing betweenthe ‘‘pterygoideus’’ and ‘‘internus’’ (Fig. 6D), andthe same idea can be applied to noncontiguous sec-tions of muscle block in other taxa. This concept ofthe ‘‘pterygoideus’’ muscle as a separated compo-nent of the sheet morphology addresses the issueof the homology of this muscle body, as raised byKleinteich and Haas (2007).

This hypothesis is extended across gnathostomephylogeny in Figure 7 and Table 3. Figure 7 is anoutline of proposed homologies based on taxaexamined and some literature accounts, placed ona simplified cladogram based on Hedges (2009).The polytomy turtles–Sphenodon–crocodiliansreflects recent evidence that turtles may be modi-fied diapsids or lepidosaurs (Lyson et al., 2010).Principal features are that the caudal adductormuscle (pink color code in Figs. 1, 3, 7, and 8) isrecognized in all taxa, and that the central adduc-tor evolves from a cylindrical shape (as in L. cha-lumnae and lungfish) into a folded sheet. Table 3sets out the same information with the addition ofthe derivatives of the m. constrictor dorsalis and ofMallatt’s (1996) concept of the primitive gnathos-tome jaw muscles. The central column of the table,muscles attached to the angle of the mouth, is con-sidered uncertain in their homologies within thisgroup and to either the rostral or the centraladductors of the holocephalan pattern.

The rostral adductors of holocephalans and elas-mobranchs are lost in euteleosts, but retained inbasal teleost lineages, examples of which are Poly-pterus senegalus, Amia calva, and Lepisosteus ocu-latus (Lauder, 1980). The folded sheet analogy forthe morphology of the rostral adductors can beseen in frogs, salamanders, and amniotes, with theU-shaped insertion wrapping around the caudaladductor, except in mammals, in which the caudaladductor no longer inserts on the mandible (Fig.7). In lungfish and coelacanth, the rostral adductorinserts medial to the caudal adductor. An evolu-tionary sequence lungfish–caecilian–amniote, withincreasing expanse of the rostral adductor inser-tion, could be proposed, but this would imply inde-pendent gain of the external part of the muscle(lateral limb of the U) in the frog–salamanderclade and in amniotes. It is more parsimonious,following Figure 7, to suggest that the U-shapedinsertion arose in the tetrapod stem and was lostin caecilians. This agrees with the conclusion ofBemis et al. (1983), who proposed that the lateralpart of the m. adductor mandibulae is lost in cae-cilians, to reduce the lateral diameter or need forlateral expansion of the head as an adaptation forburrowing. The function of the reduced adductor is

augmented by the novel mechanism of attachmentof m. interhyoideus posterior to an elongated retro-articular process of the mandible (Nussbaum,1983; Kleinteich and Haas, 2011). This progressionaccords with the view that a closed, stegokrotaphicskull table in caecilians is a derived condition(Frost et al., 2006), being made possible by reduc-tion of the conventional jaw adductor (Nussbaum,1983), although Carroll (2009; p 299) prefers analmost completely closed skull roof for the ancestorof caecilians, based on that state in the oldest cae-cilian fossil Eocaecilia micropodia.

A proposed plesiomorphic condition for the gna-thostome jaw adductors needs to incorporate fea-tures found in both holocephalans and elasmo-branchs. Although it is clear that holocephalansare the most basal extant lineage, and that theholocephalan genome contains major sequenceslost in sharks but retained in tetrapods (Ventakeshet al., 2007), the holocephalan cranial anatomy isderived, in that the upper jaw is fused to the neu-rocranium, and the gill slits have been covered byan opercular fold. The jaw adductors in batoids(rays and skates) and heterodontid sharks arecomplex (Allis, 1917; Kesteven, 1942–1945) andmay retain more homologs of the rostral holoce-phalan muscle group, but in general, sharks are agood basis for understanding the basic gnathos-tome plan (Mallatt, 1997). The identification of them. adductor mandibulae as a homolog of the mm.adductores arcuum branchialium had been sug-gested on the basis of its position in the branchialarch (Marion, 1905; Luther, 1914; Kesteven, 1942–1945), and Mallatt (1997) advanced this theory bymore clearly distinguishing between the constric-tor muscle group and the muscles of the branchialarch, to which the adductores arcuum belong (seeTable 3). The recognition here of the similar tendi-nous intersection in the shark m. adductor mandi-bulae and the mm. adductores arcuum branchia-lium supports this homology; a similar tendinousintersection in latter muscles can be seen in thefigures of Allis (1917) in Scyllium sp. and Acan-thias sp., and also in the skate Raja sp. (Marion,1905; Allis, 1917).

The hypothesis offered here (Fig. 7) is quite dif-ferent to any previous account, most obviously inthe suggestion that the main jaw adductors in tet-rapods are not homologous with those of teleosts.This has been assumed without question sinceearly accounts (Vetter, 1874). Edgeworth (1935)believed that dipnoans were plesiomorphic amonggnathostomes, and this affected all his homologyconclusions; also, Edgeworth did not accept homol-ogy between the m. adductor mandibulae articula-ris of anurans and salamanders and the m. adduc-tor mandibulae posterior of amniotes. Lightoller(1939) took the jaw adductors in all gnathostomesto be homologous most peripheral part of the su-perficial constrictor. Diogo et al. (2008) presented a

1506 P. JOHNSTON


comprehensive system based on homology of tetra-pod jaw adductors with components of the teleostjaw adductors, after the system of Vetter (1878) aselaborated by Winterbottom (1974), in which partsof the muscle are named A1, A2, and A3, from su-perficial to deep. A particular difference is themedial parts of the jaw adductor in coelacanth andlungfish (Fig. 6C,D), which are designated here asthe rostral adductor of the plesiomorphic pattern,but as the homolog of the deepest, A3 componentof the teleost pattern by Diogo et al. (2008) andDiogo and Abdala (2011). Winterbottom (1974)thought that homology between the rostral adduc-tors of chondrichthyans and teleosts such as Poly-pterus and Amia (see Table 3 and Fig. 7) was to be‘‘viewed with deep suspicion.’’ In the evolutionarysequence of muscle insertions in Figure 7, how-ever, it is reasonable to deduce this homology.

The current proposal does agree with some ele-ments of previous work, particularly with Mallatt(1996, 1997) in his recognition of the superficialand rostral muscles of the hypothetical primitivegnathostome, and their distinction from themuscles of the branchial arches. Anderson (2008)defined rostral and caudal jaw adductors in fishlineages in a similar way to that proposed here(Table 3), except that he did not recognize themedial part of the adductor group in L. chalum-nae as part of the rostral adductor. Lauder (1980)considered that the rostal adductor muscle grouphad been lost in the transition from primitive toadvanced teleosts. The salmonid used here as anexample of a teleost has the simplest jaw adduc-tor of euteleosts with an undivided m. adductormandibulae, and up to 10 different muscles inthese fish are described in the most complex cases(Friel and Wainwright, 1997). Evolutionarysequences explaining increasing complexity aredescribed by Gosline (1989), and it is concludedfrom the available data that all these muscles areall homologs of the caudal adductor sensu Table3.

The principal criterion of homology used here,continuity of topology in an evolutionary series,has not been applied in a formal cladistic manner,but is similar to the criterion of congruenceadvanced by de Pinna (1991): the hypothesis ofrostral and caudal adductors is congruent with thecladogram of Figure 7 in the sense that parsimonyis maximized on informal analysis, with onechange of state—loss of the rostral adductor in tel-eosts. Edgeworth’s (1935) view of homology amongthese muscles on this cladogram would require atleast five changes. Diogo and Abdala (2010) recog-nize many of the same elements as described here,but as different homologs, and are less specific intheir treatment of chondrichthyans and Lissam-phibians; a number of changes of state may beneeded to accommodate the Figure 7 cladogram intheir system.

This hypothesis of separate rostral and caudaladductors has been derived here on morphologicgrounds, to explain Lissamphibian jaw adductoranatomy in the context of gnathostome phylogeny,but it is also compatible with the domain ofexpression of the homeobox gene engrailed 2, asfar as data are available. The engrailed family ofgenes was first recognized in arthropods and sub-sequently found in a wide variety of bilaterians,including chordates (Holland et al., 1997). Thesegenes code for homeodomain-containing transcrip-tion factors and are thought to be important inspatial development. In the developing chordatehead, engrailed 2 is typically found at the mid-brain–hindbrain boundary, and in some of thetrigeminal muscle group (Davis et al., 1991).Engrailed 2 expression has previously not beenconsidered be indicative of homology among thetrigeminal muscles of gnathostomes (Degenhardtand Sassoon, 2001; Knight et al., 2008; Diogo andAbdala, 2010) for two main reasons: first, the jawadductor of zebra fish do not express this geneproduct (Hatta et al., 1990; Knight et al., 2008),but those of tetrapods do (Davis et al., 1991; Hem-mati-Brivanlou et al., 1991; Gardner and Barald,1992), and second, the trigeminal intermandibularmuscles, which develop from the same mesodermalanlage as the jaw adductor in all groups, do not.The first of these barriers disappears with thepresent hypothesis, in which teleost and tetrapodadductors are not homologous. The second issue isclarified by studies showing other differencesbetween intermandibular and adductor muscles:Lu et al. (2002) have demonstrated that the inter-mandibular muscles are not dependent on themyogenic regulatory factors MyoR and capsulin,unlike the jaw adductors, and Nathan et al. (2008)found that the intermandibular muscles have themolecular signature of origin from splanchnic mes-oderm, rather than from the cranial paraxial mes-oderm from which the jaw adductors arise. Sharedmesodermal origin does not necessarily meanhomology (Scholtz, 2005): in another example, theextraocular muscle m. rectus externus usuallyarises from the prechordal mesoderm, but in cer-tain taxa arises from the mandibular muscleanlage and is otherwise homologous in everyrespect (Edgeworth, 1935; Tzahor, 2009). Occur-rence of engrailed expression across the gnathos-tome taxa considered is shown in Table 3.

The limitation of the congruence between thepresent hypothesis and engrailed 2 expression isthat studies on tetrapods have not distinguishedbetween the m. adductor mandibulae posterior andthe rest of the adductor mass in frogs (Hemmati-Brivanlou et al., 1991) and chickens (Gardner andBarald, 1992): the jaw adductor is treated as a sin-gle entity, and in chickens the m. protractor ptery-goidei et quadrati, the derivative of the plesiomor-phic m. constrictor dorsalis, is not identified. In



the mouse, however, the postulated homologs of thecaudal adductor, the mm. tensor tympani and ten-sor palati, do not express this gene, whereas therest of the adductor does (Degenhardt et al., 2002).In a similar way, developmental studies with fatemapping of cranial mesoderm have not recognizedthe caudal adductor (m. adductor mandibulae pos-terior in the chicken) as a separate muscle (Evansand Noden, 2006), and support for this theory can-not be found in such work. Holland et al. (1993)showed that in the lamprey, engrailed homologs areexpressed in a particular territory of the cranialparaxial mesoderm, and predicted this would be rel-evant to gnathostomes. If the theory advanced hereis supported by future work, this prediction wouldbe confirmed. The relationship of gene expressionand morphological homology is complex and maynot be congruent (Nielsen and Martinez, 2003), butwhen it is congruent, gene expression can be a use-ful tool and could possibly be used as characters inparsimony analysis (Svensson, 2004).

Identification of a folded sheet morphology inthe main adductor mass raises the possibility of itscomponents being defined by a Cartesian pattern(i.e., referable to x, y, and z axes) of differentialgene expression, as suggested in other contexts byKuraku et al. (2010); at present such gene expres-sion has not been identified in these muscles inany gnathostome group.

The evolutionary fate of the rostral muscles andcartilages of holocephalans is uncertain; these aregenerally understood to have disappeared (Table 3),but there are some similarities with the rostralstructures of anuran larvae that have not beenexplored since Vetter (1874) and Huxley (1876). Thesuprarostral and infrarostral cartilages of anuranshave been assumed to be neomorphs (Svensson andHaas, 2005). New molecular markers may be ableto shed light on this in the future.

Lissamphian larval material has not been stud-ied here, but examination of diagrams of larvalmorphology (Cannatella, 1999; Haas, 2001) sug-gests that a folded (or twisted) sheet morphologyof the mandibular adductors could also be appliedto anuran larvae, with the difference that in lar-vae the topology is rotated internally by 908, con-sistent with the transverse rather than longitudi-nal orientation of Meckel’s cartilage in tadpoles.

The reasons for the variation in nerve coursethrough the muscle block (or sheet) are not yetcharacterized but would appear to be related to de-velopmental differences in topology and in timingof the intersection of mandibular muscle develop-ment (ventral to dorsal growth) in relation to thedevelopment of the major nerve trunks (medial tolateral growth). Such a heterochronic shift intiming is a phenomenon that is well established inAnura (Mitgutsch et al., 2008). Topologic factors,which may lead to different nerve–muscle rela-tions, are not as yet defined, but the transverse

orientation of Meckel’s cartilage and adductormuscle insertion on the suprarostral cartilage inanuran larvae are factors, which may possibly con-tribute.

M. Levator Anguli Oris

The identification of m. levator anguli oris in A.truei is the first record of such a muscle in a frog.Superficial muscles attached to the labial carti-lages, rictal plate, or their equivalent at the cornerof the mouth are well known in chondrichthyanand osteichthyan fish, dipnoans, and lepidosaurianreptiles (Edgeworth, 1935; Haas, 1973; Mallatt,1996). In Lissamphibia, the only account of such amuscle is that of Luther (1914), who notes severalmorphologies in salamanders (p 52):

‘‘Generally the most rostral portion of the m. adduc-tor mandibulae externus, immediately dorsal to itsinsertion, is more or less firmly connected to the skinat the corner of the mouth. Frequently this attach-ment is so close that it is difficult to separate the skinwithout damaging it. In Necturus (his Fig. 49) itseems to me that a body of fibers inserts directly ontothe skin at the corner of the mouth. This situationleads to that shown by Siren (his Fig. 38). Here a ven-tral portion of m. adductor mandibulae externus arisesby means of a broad tendinous sheet from the caudalborder of the dentary and passes rostrally to attach tothe upper lip (rao). This can on account of its functionbe designated as m. retractor anguli oris.’’

[my translation]

In reptiles the ‘‘anguli oris’’ muscles have beeninterpreted as the most superficial section of them. adductor mandibulae (Haas, 1973; Diogo et al.,2008). The origin of m. levator anguli oris in A.truei is not, however, part of the externus compo-nent, but has a separate origin on the roof of themouth. The typical site of origin of the anguli orismuscles in lepidosaurs on the postorbital and baris missing from the anuran skull, but the similar-ity of insertion suggests homology with the lepido-saur muscle. The work of Mallatt (1996) suggeststhat, phylogenetically, the corner of the mouthmay be a remnant of the plesiomorphic mouthopening of agnathans; if this is so, the ‘‘angulioris’’ muscles could be homologs of the most rostraltrigeminal muscle of agnathans and holocepha-lans. The presence of muscles attached to therictal plate of some lungfish and salamandersleads to the m. levator anguli oris of A. truei beingrecognized as a plesiomorphic state in Table 2.

Mm. Depressor Mandibulae and LevatorBulbi

The morphology of m. depressor mandibulae inL. hochstetteri and A. truei shows the two majorcomponents commonly found in Lissamphibia(Starrett, 1968; Bauer, 1997); the absence of a tym-

1508 P. JOHNSTON


panic ring in these two frogs results in a simplerarrangement than in other frogs. According toLightoller (1939), the plesiomorphic situation ofthe depressor is for the cervical component toinsert laterally, as in L. hochstetteri, with the op-posite finding in A. truei being a derived state.

Luther (1914) described the caudal attachmentof m. levator bulbi to the tip of the processus zygo-maticus of the squamosal in the anurans he exam-ined, whereas in the species examined here, abroader insertion is noted; this does not seem tohave been recorded for any other frogs or for sala-manders, and this may be a synapomorphy for aLeiopelma–Ascaphus clade. The separate trans-verse component of m. levator bulbi identified byLuther is not found in L. hochstetteri and A. truei.

Tongue, Hyobranchial, and LaryngealMuscles

The tongues of these two frogs are similar ingross morphology, with tongues attached to the oralmucosa around their circumference, thus without afree edge—the ‘‘discoglossoid’’ tongue morphology ofMagimel-Pelonnier (1924). Such tongues are capa-ble of only limited protrusion from the mouth(Regal and Gans, 1976; Nishikawa and Cannatella,1991), although no corresponding features in themyology have been noted. The findings presentedhere contradict the suggestion of Horton (1982)that the number of interdigitations of mm. hyoglos-sus and genioglossus are useful as phylogeneticcharacters: three or four digitations are found herein L. hochstetteri, and three in A. truei, whereasHorton (1982) found two and four, respectively, inher specimens. In both L. hochstetteri and A. truei,the medial component of the geniohyoideus as seenin other frogs (Trewavas, 1933) is absent; Trewavasdescribed a single lateral belly of m. geniohyoideusin her Leiopelma specimen, whereas two adjacentparts of the lateral muscle have been recognizedhere. A medial insertion of m. geniohyoideus wasshown to be the plesiomorphic state by Jarvik(1963), and its loss is a potential synapomorphy forLeiopelma and Ascaphus. Cannatella (1985) identi-fied an absence of longitudinal overlap in theattachments of mm. geniohyoideus and sternohyoi-deus on the branchial plate as a synapomorphy ofLeiopelma and Ascaphus, but this was not convinc-ing in the material examined here.

Other features such as the absence of m. omo-hyoideus, the fusion of m. petrohyoideus IV withthe ventral part of m. constrictor laryngis in A.truei, and the dorsal origin of m. constrictor lar-yngis in L. hochstetteri are known in various otherfrogs, and are distributed in anurans withoutobvious phylogenetic significance (Trewavas, 1933).

The muscles of the larynx in A. truei are simplerthan those of any other anuran documented, inthat the m. constrictor laryngis comprises a single

section, without the additional componentsreferred to as mm. constrictor externus, anterior,posterior, dorsalis, and ventralis sensu Trewavas(1933). This may be related to the absence of vocalfunction in A. truei, although the presence of vocalchords in this frog, which does not appear to havebeen documented previously, suggests that thissilence is a derived state. The Ascaphus laryngealmusculature is evidently more advanced than thatof salamanders, in which the m. dilator laryngisarises from the dorsolateral fascia of the neck ad-jacent to the m. cucullaris (Hilton, 1952), which isconsistent with the plesiomorphic situation (Edge-worth, 1935; Ericsson et al., 2011). In L. hochstet-teri, the m. constrictor laryngis ventralis asdescribed here extends further rostral than thedepiction of Trewavas (1933) of L. archeyi andinserts on a midline raphe in the space betweenthe cricoid and branchial cartilages, similar to thatseen in A. truei.

Green (1988) cited Trewavas (1933) for his state-ment that Leiopelma has ‘‘no true voice-box,’’ but Ido not read Trewavas this way, and the findingshere in L. hochetteri are of a larynx similar to thatof other anurans.

There is a separate comparative informationpublished on both laryngeal morphology and vocalfunction in anurans (Wells, 2007), but an attemptto correlate these data across a range of taxa doesnot appear to have been undertaken; this limitsphylogenetic and functional inference from theirlaryngeal morphology. The laryngeal muscles inAnura are complex, as is their terminology, whichmay benefit from revision.

CONCLUSIONS

Synapomorphies can only be named in the con-text of a phylogeny including those characters; twopotential synapomorphies for a Leiopelma–Asca-phus clade are identified here: the loss of the ple-siomorphic medial component of m. geniohyoideus,and the presence of a caudal attachment of m.levator bulbi to a ligamentous thickening of thefascia covering the adductor muscles. Publishedmorphological character sets for adult Anura arerelatively small (Frost et al., 2007); the recent con-gruent molecular phylogenies (see ‘‘Introduction’’section) with strong support for a Leiopelma–Asca-phus clade make the identification of plesiomor-phic and synapomorphic features in these generaimportant for understanding their morphologicalevolution. More morphological characters areneeded, and a number of new ones are suggestedhere.

A new hypothesis is advanced here for thehomology of the jaw adductors in Lissamphibiaand developed across gnathostome phylogeny:this is based on holocephalan and elasmobranchmorphology as representing the plesiomorphic



state and is influenced by the work of Mallatt(1996, 1997) in his concept of the primitive gna-thostome morphology. This differs from previoustheories, which are based on dipnoans (Edgeworth,1935) and teleosts (Diogo et al., 2008) as models.The system outlined here applies to all gnathos-tome groups, and if supported by further studiescould form the basis for a unified nomenclature.The possibility that expression of engrailed 2 isuseful evidence for investigations of homologyopens the way to more detailed studies of thisgene in chondrichthyan fish and tetrapods. Theevolutionary history of the rostral trigeminalmuscles of holocephalans also awaits resolution,and this may be achieved when suitable molecularmarkers are available.

ACKNOWLEDGMENTS

For access to specimens, the author thanksBrian Gill and Tom Trnski (Auckland Museum),Andrew Stewart (Museum of New Zealand, TePapa Tongarewa), and Jean Joss (Macquarie Uni-versity). He is grateful to Jessie Maisano, Univer-sity of Texas Digital Morphology Group for the CTscan data and to Larry Franks and Rachel Ber-quist, Digital Fish Library, University of Califor-nia, San Diego for the MRI scans. The Digital FishLibrary project is funded by NSF grant numberDBI-0446389. Christine Lorre kindly helped withacquiring literature. He thanks Rolf Ericsson foran early copy of his manuscript, and two anony-mous reviewers whose comments have improvedthis paper.

LITERATURE CITED

Allis EP. 1897. The cranial muscles and cranial and first spinalnerves in Amia calva. J Morphol 12:487–772.

Allis EP. 1917. The homologies of the muscles related to the vis-ceral arches of the gnathostome fishes. Q J Microsc Sci62:303–406.

Allis EP. 1922. The cranial anatomy of Polypterus, with specialreference to Polypterus bichir. J Anat 56:180–294.

Anderson PSL. 2008. Cranial muscle homology across moderngnathostomes. Biol J Linn Soc 94:195–216.

Bartsch P. 1994. Development of the cranium of Neoceratodusforsteri, with a discussion of the suspensorium and the oper-cular apparatus in Dipnoi. Zoomorphology 114:1–31.

Bauer W. 1997. A contribution to the morphology of visceraljaw-opening muscles of urodeles (Amphibia: Caudata). J Mor-phol 233:77–97.

Bemis WE, Schwenk K, Wake MH. 1983. Morphology and func-tion of the feeding apparatus in Dermophis mexicanus. Zool JLinn Soc 77:75–96.

Bossuyt F, Roelants K. 2009. Frogs and toads (Anura). In:Hedges SB, Kumar S, editors. The Timetree of Life. Oxford:Oxford University Press. pp 357–364.

Cannatella DC. 1985. A Phylogeny of Primitive Frogs (Archaeo-batracians) [PhD]. Lawrence, Kansas: University of Kansas.

Cannatella DC. 1999. Architecture: Cranial and axial musculo-skeleton. In: McDiarmid RW, Altig R, editors. Tadpoles:The Biology of Anuran Larvae. Chicago: University of Chi-cago Press. pp 52–91.

Carroll RL. 2007. The Palaeozoic ancestry of salamanders, frogsand caecilians. Zool J Linn Soc 150 (Supp l 1):1–140.

Carroll RL. 2009. The Rise of Amphibians: 365 Million Years ofEvolution. Baltimore: Johns Hopkins University Press.

Carroll RL, Holmes R. 1980. The skull and jaw musculature asguides to the ancestry of salamanders. Zool J Linn Soc 68:1–40.

Compagno LJV. 1988. Sharks of the Order Carchariniformes.Princeton: Princeton University Press.

Davis CA, Holmyard DP, Millen KJ, Joyner AL. 1991. Examin-ing pattern formation in mouse, chicken and frog embryoswith an En-specific antiserum. Development 111:287–298.

de Pinna MCC. 1991. Concepts and tests of homology in thecladistic paradigm. Cladistics 7:367–394.

Degenhardt K, Sassoon DA. 2001. A role for Engrailed-2 indetermination of skeletal muscle physiologic properties. DevBiol 231:175–189.

Degenhardt K, Rentschler S, Fishmann G, Sassoon DA. 2002.Cellular and cis-regulation of En-2 expression in the mandib-ular arch. Mech Dev 111:125–136.

Didier DA. 1995. Phylogenetic systematics of extant chimeroidfishes (Holocephali, Chimaeroidei). Am Mus Novit 3119:1–86.

Diogo R, Abdala V. 2010. Muscles of Vertebrates. Enfield, NewHampshire: CRC Press.482p.

Diogo R, Abdala V. 2011. The head muscles of dipnoans—Areview on the homologies and evolution of these muscleswithin vertebrates. In: Jorgensen JM, Joss J, editors. TheBiology of Lungfishes. Enfield, New Hampshire: CRC Press.pp 169–218.

Diogo R, Abdala V, Lonergan N, Wood B. 2008. From fish tomodern humans—Comparative anatomy, homologies andevolution of the head and neck musculature. J Anat 213:391–424.

Duellman WE, Trueb L. 1986. Biology of the Amphibia. NewYork: McGraw-Hill.

Edgeworth FH. 1935. The Cranial Muscles of Vertebrates. Cam-bridge: Cambridge University Press.

Ericsson R, Joss J, Olsson L. 2011. Early head development inthe Australian lungfish. In: Jorgensen JM, Joss J, editors.The Biology of Lungfishes. Enfield, New Hampshirew: CRCPress. pp 149–168.

Estes R, Reig OA. 1973. The early fossil record of frogs: Areview of the evidence. In: Vial JL, editor. Evolutionary Biol-ogy of the Anurans. Columbia: University of Missouri Press.pp 11–64.

Evans DJR, Noden DM. 2006. Spatial relations between aviancraniofacial neural crest and paraxial mesoderm cells. DevDyn 235:1310–1325.

Ford LS, Cannatella DC. 1993. The major clades of frogs.Herpetol Monogr 7:94–117.

Forey P. 1998. A History of the Coelacanth Fishes. London:Chapman & Hall.

Frazier M. 1924. A contribution to the anatomy of the amphib-ian larynx. J Morphol 39:285–293.

Friel JP, Wainwright PC. 1997. A model system of structuralduplication: Homologies of the adductor mandibulae musclesin tetraodontiform fishes. Syst Biol 46:441–463.

Frost DR, Grant T, Faivovich J, Bain RH, Haas A, HaddadCFB, De Sa RO, Channing A, Wilkinson M, Donnellan SC,Raxworthy CJ, Campbell JA, Blotto BL, Moler P, Drewes RC,Nussbaum RA, Lynch JD, Green DM, Wheeler WC. 2006. Theamphibian tree of life. Bull Am Mus Nat Hist 297:1–370.

Frost DR, Grant T, Faivovich J, Bain RH, Haas A, HaddadCFB, De Sa RO, Channing A, Wilkinson M, Donnellan SC,Raxworthy CJ, Campbell JA, Blotto BL, Moler P, Drewes RC,Nussbaum RA, Lynch JD, Green DM, Wheeler WC. 2007. IsThe Amphibian Tree of Life really fatally flawed? Cladistics23:1–11.

Gao K-Q, Wang Y. 2001. Mesozoic anurans from Liaoning prov-ince, China, and phylogenetic relationships of archaeobatra-chian anuran clades. J Vertebr Paleontol 21:460–476.

1510 P. JOHNSTON


Gardner CA, Barald KF. 1992. Expression patterns ofengrailed-like proteins in the chick embryo. Dev Dyn193:370–388.

Gaupp E. 1896. A. Ecker’s und R.Wiedersheim’s Anatomie desFrosches. Braunschweg: Friedrich Vieweg und Sohn.

Gosline WA. 1989. Two patterns of differentiation in the jawmuscles of teleostean fishes. J Zool 218:649–661.

Green DM. 1988. Antipredator behaviour and skin glands inthe New Zealand native frogs, genus Leiopelma. N Z J Zool15:39–45.

Grogan ED, Lund R, Didier D. 1999. Description of the chi-maerid jaw and its phylogenetic origins. J Morphol 239:45–59.

Haas A. 2001. Mandibular arch musculature of anuran tad-poles, with comments on homologies of amphibian jawmuscles. J Morphol 247:1–33.

Haas G. 1973. Muscles of the jaws and associated structures inthe Rhynchocephalia and Squamata. In: Gans C, Parsons TS,editors. Biology of the Reptilia. New York: Academic Press.pp 285–490.

Hatta K, Schilling TF, BreMiller RA, Kimmel CB. 1990. Specifi-cation of jaw muscle identity in zebrafish: Correlation withengrailed-homeoprotein expression. Science 250:802–805.

Hedges SB. 2009. Vertebrates (Vertebrata). In: Hedges SB,Kumar S, editors. The Timetree of Life. Oxford: Oxford Uni-versity Press. pp 309–314.

Hemmati-Brivanlou A, de la Torre JR, Holt C, Harland RM.1991. Cephalic expression and molecular characterization ofXenopus En-2. Development 111:715–724.

Hermida GN, Farias A. 2009. Morphology and histology of thelarynx of the common toad Rhinella arenarum (Hensel, 1867)(Anura, Bufonidae). Acta Zool (Stockholm) 90:326–338.

Hilton WA. 1952. The pulmonary respiratory system of sala-manders. Herpetologica 8:87–92.

Holland ND, Holland LZ, Honma Y, Fuji T. 1993. Engrailedexpression during development of a lamprey, Lampetra japon-ica: A possible clue to the homologies between agnathan andgnathostome muscles of the mandibular arch. Dev GrowthDiffer 35:153–160.

Holland LZ, Kene M, Williams NA, Holland ND. 1997.Sequence and embryonic expression of the amphioxusengrailed gene (AmphiEn): The metameric pattern of tran-scription resembles that of its segment-polarity homolog inDrosophila. Development 124:1723–1732.

Holliday CM. 2009. New insights into dinosaur jaw muscleanatomy. Anat Rec 292:1246–1265.

Holliday CM, Witmer LM. 2007. Archosaur adductor chamberevolution: Integration of musculoskeletal and topological cri-teria in jaw muscle homology. J Morphol 268:457–484.

Horton P. 1982. Diversity and systematic significance of anurantongue musculature. Copeia 1982:595–602.

Huxley TH. 1876. On the nature of the craniofacial app aratusof Petromyzon. J Anat Physiol 10:412–429.

Iordansky NN. 1992. Jaw muscles of the Urodela and Anura:Some features of development, functions and homology. ZoolJahrb Abt Anat 122:225–232.

Iordansky NN. 1996. Evolution of the musculature of the jawapp aratus in the amphibia. Adv Amphibian Res FormerSoviet Union 1:3–26.

Irisarri I, San Mauro D, Green DM, Zardoya R. 2010. The com-plete mitochondrial genome of the relict frog Leiopelmaarcheyi: Insights into the root of the frog tree of life. Mito-chondr DNA 21:173–182.

Jarvik E. 1963. The composition of the intermandibular divisionof the head in fish and tetrapods and the diphyletic originof the tetrapod tongue. K Svenska Vetenskapsakad Handl9:1–74.

Kesteven HL. 1942–1945. The evolution of the skull and the ce-phalic muscles: A comparative study of their developmentand adult morphology. Aust Mus Mem 8:1–361.

Kleinteich T, Haas A. 2007. Cranial musculature in the larva ofthe caecilian, Ichthyophis kohtaoensis (Lissamphibia: Gymno-phiona). J Morphol 268:74–88.

Kleinteich T, Haas A. 2011. The hyal and ventral branchialmuscles in caecilian and salamander larvae: Homologies andevolution. J Morphol 272:598–613

Knight RD, Mebus K, Roehl HH. 2008. Mandibular arch muscleidentity is regulated by a conserved molecular process duringvertebrate development. J Exp Zool 310B:355–369.

Kuraku S, Takio Y, Suguhara F, Takechi M, Kuratani S. 2010.Evolution of oropharyngeal patterning mechnisms involvingDlx and endothelins in vertebrates. Dev Biol 341:315–323.

Lauder GV. 1980. Evolution of the feeding mechanism in primi-tive actinopterygian fishes: A functional anatomical analysisof Polypterus, Lepisosteus and Amia. J Morphol 163:283–317.

Lightoller GHS. 1939. Probable homologues. A study of thecomparative anatomy of the mandibular and hyoid archesand their musculature. Part I. Comparative myology. TransZool Soc Lond 24:349–382.

Lu JR, Bassel-Duby R, Hawkins A, Chang P, Valdez R, Wu H,Gan L, Shelton JM, Richardson JA, Olson EN. 2002. Controlof facial muscle development by MyoR and capsulin. Science298:2378–2381.

Lubosch W. 1914. Vergleichende Anatomie der Kaumuskulaturder Wirbiltiere, in funf Teilen: 1. Die Kaumuskulatur derAmphibien. Jena Z Naturwiss 53:51–188.

Lubosch W. 1938. Muskeln des Kopfes. C. Amphibien und Sau-rien. In: Bolk L, Gopp ert E, Kallius E, Lubosch W, editors.Handbuch der vergleichenden Anatomie der Wirbiltiere. Ber-lin, Wien: Urban & Schwarzenberg. pp 1025–1064.

Luther A. 1909. Untersuchungen uber die vom N. trigeminusinnervierte Muskulatur der Selachier (Haie und Rochen).Acta Soc Sci Fenn 26:1–176.

Luther A. 1914. Uber die vom N. trigeminus versorgte Musku-latur der Amphibien, mit einem vergleichenden Ausblick uberden Adductor mandibulae der Gnathostomen, und einemBeitrag zum Verstandnis der Organisation der Anurenlarven.Acta Soc Sci Fenn 44:1–151.

Lynch JD. 1973. The transition from archaic to advanced frogs.In: Vial JL, editor. Evolutionary Biology of the Anurans. Co-lumbia: University of Missouri Press. pp 133–182.

Lyson TR, Bever GS, Bhullar B-AS, Joyce WG, Gauthier JA.2010. Transitional fossils and the origin of turtles. Biol Lett6:830–833.

Magimel-Pelonnier OL. 1924. La langue des amphibiens. These,Faculte des Sciences de Paris. Bordeaux: A. Saugnac and E.Provillard. 260 p.

Mallatt J. 1996. Ventilation and the origin of jawed vertebrates:A new mouth. Zool J Linn Soc 117:329–404.

Mallatt J. 1997. Shark pharyngeal muscles and early vertebrateevolution. Acta Zool (Stockholm) 78:279–294.

Marion GE. 1905. Mandibular and pharyngeal muscles of Acan-thias and Raja. Am Nat 39:891–924.

Millot J, Anthony J. 1958. Anatomie de Latimeria chalumnae—I, squelette, muscles et formation de soutiens. Paris: CNRS.

Mitgutsch C, Piekariski N, Olsson L, Haas A. 2008. Hetero-chronic shifts during early cranial neural crest cell migrationin two ranid frogs. Acta Zool (Stockholm) 89:69–78.

Nathan E, Monovich A, Tirosh-Finkel L, Harrelson Z, Rousso T,Rinon A, Harel I, Evans SM, Tzahor E. 2008. The contribu-tion of Islet 1-expressing splanchnic mesoderm cells to dis-tinct branchiomeric muscles reveals significant heterogeneityin head muscle development. Development 135:647–657.

Nielsen C, Martinez P. 2003. Patterns of gene expression:Homology or homocracy. Dev Genes Evol 213:149–154.

Nishikawa KC, Cannatella DC. 1991. Kinematics of prey cap-ture in the tailed frog Ascaphus truei (Anura: Ascaphidae).Zool J Linn Soc 103:289–307.

Nussbaum RA. 1983. The evolution of an unique dual jaw-closing mechanism in caecilians (Amphibia: Gymnophiona)and its bearing on caecilian ancestry. J Zool 199:545–554.

Pugener LA, Maglia A, Trueb L. 2003. Revisiting the contribu-tion of larval characters to an analysis of phylogenetic analy-sis of basal anurans. Zool J Linn Soc 139:129–155.

Regal PJ, Gans C. 1976. Functional aspects of the evolution ofthe tongue in frogs. Evolution 30:718–734.



Ritland RM. 1955. Studies on the post-cranial morphology ofAscaphus truei. II. Myology. J Morphol 97:215–282.

Rosen DE, Patterson C. 1969. The structure and relationshipsof the paracanthopterygian fishes. Bull Am Mus Nat Hist141:357–474.

Save-Soderbergh G. 1945. Notes on the trigeminal musculaturein non-mammalian tetrapods. Nova Acta Reg Soc Sci Upsal13:1–50.

Schlosser G, Roth G. 1995. Distribution of cranial and rostralspinal nerves in tadpoles of the frog Discoglossus pictus(Discoglossidae). J Morphol 226:189–212.

Scholtz G. 2005. Homology and ontogeny: Pattern and process incomparative developmental biology. Theory Biosci 124:121–143.

Starrett PH. 1968. The phylogenetic significance of the jawmusculature in anuran amphibians [PhD]. Ann Arbor: Uni-versity of Michigan. 179 p.

Stephenson EM. 1951. The anatomy of the head of the NewZealand frog, Leiopelma. Trans Zool Soc Lond 27:255–305.

Svensson ME. 2004. Homology and homocracy revisited: Geneexpression patterns and hypotheses of homology. Dev GenesEvol 214:418–421.

Svensson ME, Haas A. 2005. Evolutionary innovation in thevertebrate jaw: A derived morphology in anuran tadpoles andits possible developmental origin. Bioessays 27:526–532.

Trewavas E. 1933. The hyoid and larynx of the Anura. PhilosTrans Roy Soc Lond Ser B 222:401–527.

Trueb L. 1993. Patterns of cranial diversity among the Lissam-phibia. In: Hanken J, Hall BK, editors. The Skull. Chicagoand London: University of Chicago Press. pp 255–343.

Tzahor E. 2009. Heart and craniofacial muscle development: Anew developmental theme of distinct myogenic fields. DevBiol 327:273–279.

Ventakesh B, Kirkness EF, Loh Y-H, Halpern AL, Lee AP, John-son J, Dandona N, Viswanathan LD, Tay A, Venter JC,Strausberg RL, Brenner S. 2007. Survey sequencing and com-parative analysis of the elephant shark (Callorhinchus milli)genome. PLoS Biol 5:932–944.

Vetter B. 1874. Untersuchungen zur vergleichenden Anatomieder Kiemen- und Kiefermuskulatur der Fische. Jena ZNaturwiss 8:405–458.

Vetter B. 1878. Untersuchungen zur vergleichenden Anatomieder Kiemen- und Kiefemuskulatur der Fische. Jena ZNaturwiss 12:431–550.

Wagner DS. 1934. Liopelma studies Nr. 2. On the cranial char-acters of Liopelma hochstetteri. Anat Anz 79:65–77.

Wells KD. 2007. The Ecology and Behaviour of Amphibians.Chicago and London: University of Chicago Press.

Wiens JJ. 2011. Re-evolution of lost mandibular teeth in frogsafter more than 200 million years, and re-evaluating Dollo’slaw. Evolution 65:1283–1296.

Williams RRG. 1997. Bones and muscles of the suspensorium inthe galaxioids and Lepidogalaxias salamandroides (Teleostei:Osmeriformes) and their phylogenetic significance. Rec AustMus 49:139–166.

Winterbottom R. 1974. A descriptive synonymy of the striatedmuscles of the Teleostei. Proc Acad Nat Sci Philad 125:225–317.

Ziermann JM, Olsson L. 2007. Patterns of spatial and temporalcranial muscle development in the African clawed frog Xeno-pus laevis (Anura: Pipidae). J Morphol 268:791–801.

1512 P. JOHNSTON


cranial muscles of the anurans leiopelma hochstetteri and ascaphus truei and the homologies of the...

Documents