Myological variability in a decoupled skeletal system: Batoid cranial anatomyutweb.ut.edu/hosted/faculty/dhuber/images/Kolmann et al... · 2016-03-03 · Myological Variability in

Myological Variability in a Decoupled Skeletal System:Batoid Cranial Anatomy

Matthew A. Kolmann,1* Daniel R. Huber,2 Mason N. Dean,3 and R. Dean Grubbs1

1Florida State University Coastal and Marine Laboratory, St. Teresa, Florida 323582Department of Biology, University of Tampa, Tampa, Florida 336063Department of Biomaterials, Max Planck Institute of Colloids and Interfaces, Potsdam 14424, Germany

ABSTRACT Chondrichthyans (sharks, batoids, andchimaeras) have simple feeding mechanisms owing totheir relatively few cranial skeletal elements. However,the indirect association of the jaws to the cranium(euhyostylic jaw suspension) has resulted in myriadcranial muscle rearrangements of both the hyoid andmandibular elements. We examined the cranial muscu-lature of an abbreviated phylogenetic representation ofbatoid fishes, including skates, guitarfishes and with aparticular focus on stingrays. We identified homologousmuscle groups across these taxa and describe changesin gross morphology across developmental and func-tional muscle groups, with the goal of exploring howdecoupling of the jaws from the skull has effected mus-cular arrangement. In particular, we focus on the cra-nial anatomy of durophagous and nondurophagousbatoids, as the former display marked differences inmorphology compared to the latter. Durophagous sting-rays are characterized by hypertrophied jaw adductors,reliance on pennate versus fusiform muscle fiber archi-tecture, tendinous rather than aponeurotic muscleinsertions, and an overall reduction in mandibularkinesis. Nondurophagous stingrays have muscles thatrely on aponeurotic insertions onto the skeletal struc-ture, and display musculoskeletal specialization for jawprotrusion and independent lower jaw kinesis, relativeto durophagous stingrays. We find that among extantchondrichthyans, considerable variation exists in thehyoid and mandibular muscles, slightly less so inhypaxial muscles, whereas branchial muscles are over-whelmingly conserved. As chondrichthyans occupy aposition sister to all other living gnathostomes, ourunderstanding of the structure and function of earlyvertebrate feeding systems rests heavily on under-standing chondrichthyan cranial anatomy. Our findingshighlight the incredible variation in muscular complex-ity across chondrichthyans in general and batoids inparticular. J. Morphol. 000:000–000, 2014. VC 2014 Wiley

Periodicals, Inc.

KEY WORDS: Myliobatidae; batoid; gnathostome; dur-ophagy; jaw suspension

INTRODUCTION

The batoid fishes (stingrays, electric rays,skates, and guitarfishes) possess arguably one ofthe most functionally autonomous feeding mecha-nism within vertebrates. Whereas most vertebratespecies show a strong association between the

upper jaws and braincase, in batoids the only skel-etal links to the upper [palatoquadrate (PQ)] andlower [Meckel’s cartilage (MK)] jaws are thepaired hyomandibular cartilages, rod-like connect-ing struts between the corners of the jaws and theotic region of the chondrocranium (Maisey, 1980;Wilga, 2002; Dean et al., 2007a; Motta and Huber,2012). Unlike sharks, this jaw suspension ofbatoids (known as euhyostyly) is solely hyoidbased, without the added assistance of ligamentsor skeletal processes to anchor the upper jaw tothe chondrocranium to limit or guide jaw move-ment (Fig. 1; Maisey, 1980; Wilga, 2002). Theresult is a feeding mechanism with a relativelysimple skeletal structure that can exhibit greatfreedom of movement (Wilga and Motta, 1998;Dean and Motta, 2004a).

Despite the relatively few elements in the feed-ing mechanism, the ranges of ecological nichesused by batoids are comparatively diverse (Deanet al., 2007a). Batoids exhibit a rich array of feed-ing modes ranging from benthic generalists to spe-cialists on shelled prey, and from predators onlarge, highly mobile midwater prey to pelagicplanktivores (Bigelow and Schroeder, 1953;Notarbartolo-di-Sciara, 1987; Dean et al., 2005,2007a; Collins et al., 2007). Although the func-tional significance of the observed interspecificvariations in hyomandibular morphology on diet is

Additional Supporting Information may be found in the online ver-sion of this article.

Contract grant sponsor: Florida State University Coastal andMarine Laboratory graduate research (M.A.K.); Contract grantsponsor: FSUCML (R.D.G.).

*Correspondence to: M.A. Kolmann; Department of Ecology andEvolutionary Biology, University of Toronto Scarborough, 1265 Mili-tary Trail, Toronto, ON M1C 1A4, Canada. E-mail:[email protected]

Received 28 July 2013; Revised 28 January 2014;Accepted 27 February 2014.

Published online 00 Month 2014 inWiley Online Library (wileyonlinelibrary.com).DOI 10.1002/jmor.20263

VC 2014 WILEY PERIODICALS, INC.

JOURNAL OF MORPHOLOGY 00:00–00 (2014)

unclear, jaw is thought to be highly reflective ofdiet (Summers, 2000; Dean et al., 2007a). Thejaws of narcinid (Narcinidae) electric rays, forexample, have flexible symphyses and a gape con-strained laterally by labial cartilages, permittingrapid jaw protrusion for suction feeding specializa-tion on vermiform prey (worms and eels; Deanand Motta, 2004a,b). Both torpedinid (Torpedini-dae) electric rays and butterfly rays (Gymnuridae)have flexible, gracile jaws with large gapes, allow-ing engulfment of food items much larger than theresting mouth opening (Dean et al., 2007a). Spe-cializations for crushing or cracking durable prey(durophagy) have evolved several times, in someguitarfishes (e.g., Rhina and Zapteryx), some basalstingrays (e.g., Pastinachus) and all members ofthe families Myliobatidae and Rhinopteridae. Theadvent of a durophagous feeding mode in thesedisparate taxa is thought to be evolutionarily inde-pendent, although durophagy in Myliobatidae andRhinopteridae is thought to have arisen oncethrough a common ancestor and then lost in theplanktivorous Mobulidae (Notarbartolo-di-Sciara,1987; Summers, 2000; Dean et al., 2007a). Mor-phological adaptations for durophagy typicallyinvolve modifications of teeth (e.g., reduction of

tooth number, flattening of teeth into pavement-like dentition) and reinforcement of the jaws viasymphyseal fusion, cortical thickening and theaddition of trabecular struts to buttress the inte-rior of the jaw and prevent collapse during thecrushing of molluscan prey (Summers, 2000; Deanet al., 2007a). We expect that myliobatid duropha-gous stingrays also share musculoskeletal traits incommon with other durophagous elasmobranchs;namely, hypertrophied muscles, rigid/reinforcedskeletons, and muscle fiber architecture favoringincreased force generation (Kolmann and Huber,2009).

The range of batoid diets and comparative sim-plicity of the skeletal system provide a fascinatingpalette for studies of evolutionary ecomorphology.In particular, the multiple evolutionary instancesof specialized diets from hypothesized generalistancestors (Dean et al., 2007a) and the presence ofdrastically different ecologies and morphologies insister clades (e.g., narcinid vs. torpedinid electricrays, durophagous myliobatid vs. filter-feedingmobulid rays) may suggest that 1) the evolution ofdietary and morphological specializations cannotsimply be predicted by phylogeny and 2) the func-tional elements within the feeding system have

Fig. 1. Methods of jaw suspension in chondrichthyan fishes. (a) Holostylic—Hydrolagus, (b)Orbitostylic—Squalus, (c) Euhyostylic—Urobatis, (d) Hyostylic with ethmoid attachments—Carcharhinus.

2 M.A. KOLMANN ET AL.

Journal of Morphology

been decoupled and can therefore evolve independ-ently. The “decoupling hypothesis” has been sug-gested for both teleosts (Schaefer and Lauder,1996; Hulsey et al., 2006; Lujan and Armbruster,2012) and elasmobranchs (Wilga and Motta, 1998;Dean et al., 2007a; Motta and Huber, 2012) andposits that as interacting skeletal elements (e.g.,portions of the jaw and hyoid arches) are not con-strained to evolve together, reconfiguration of thefeeding mechanism can lead to a wide assemblageof drastically different functions and ecologies.That the jaws enjoy great freedom in their connec-tion to the chondrocranium and likely rely heavilyon hyomandibular movement (e.g., Dean andMotta, 2004a, b) suggests that jaw and hyoid mus-culature (Fig. 2) plays a vital role in suspending,stabilizing, and actuating the feeding mechanism.Understanding variation in muscular characterstates among species will allow insight into howdietary and kinematic diversity have evolved andare achieved with a cranial skeleton with few mov-ing skeletal elements.

Recent reevaluation of batoids as sistergroup tosharks (Douady et al., 2003; Aschliman et al.,2012) necessitates equal “weight” is given to bothbatoid and selachian character states in phyloge-netic reconstructions of stem elasmobranch cranialanatomy. Although musculoskeletal form and func-tion have been investigated in a variety of nonba-toid chondrichthyans (e.g., Wilga and Motta, 2000;Huber et al., 2005, 2008; Kolmann and Huber,

2009; Mara et al., 2009), the arrangement of thecranial musculature in batoids has received scantattention from a functional standpoint (but seeWilga and Motta, 1998; Summers, 2000; Dean andMotta, 2004a; Sasko et al., 2006; Mulvany andMotta, 2013). Even fewer studies have sought toreconcile study of batoid musculature with othervertebrates (but see Miyake et al., 1992). Phyloge-nies incorporating jaw suspension as a criticalcharacter (Maisey, 1980) have recently been sup-ported by subsequent molecular studies (Douadyet al., 2003; Aschliman et al., 2012); suggestingfeeding morphology is inherently tied to the diver-sification of stem chondrichthyans lineages.Despite the considerable evolutionary time elapsedas the divergence of these stem lineages, jaw sus-pension modes remain relatively lineage specific(Maisey, 1980). That muscle morphology follows askeletal template would lead us to hypothesizethat fixation of jaw suspension modes deep withinthe phylogenetic history of chondrichthyans is mir-rored by restriction of hyoid muscle variability tothese ancient nodes as well. As a counterpoint,given the array of dietary niches occupied by mod-ern elasmobranchs (Dean et al., 2007a), we wouldexpect most modern ecomorphological variation tooccur in mandibular musculature.

Here, we describe the feeding musculatureacross stingrays in contrast to other batoids(skates and guitarfishes, Fig. 3). We then identifyhomologous muscles across elasmobranchs

Fig. 2. Hypothetical generalized muscle vectors in batoids.

3STINGRAY CRANIAL MUSCLES


(sharks 1 rays) and then chondrichthyans [ratfish-es 1 (sharks 1 rays), Aschliman et al., 2012] as awhole. Special attention is given to durophagousstingrays, as they exhibit quite divergent muscu-lar anatomy from other batoids. We then use mor-phological observations to infer the possiblefunctional implications of observed musculararrangements. When possible, we follow the termi-nology of Miyake et al. (1992), which, building onthe legacy of Edgeworth’s (1935) research intohomology of cranial musculature across verte-brates, incorporated an embryological and evolu-tionary perspective. This methodology allowsfurther comparison to other vertebrate taxa andbuilds a common etymological framework. Ourphylogenetic schema follows that of Aschlimanet al. (2012), the most comprehensive systematicstudy of batoids to date, incorporating molecular,morphological, and fossil data in understandingthe evolutionary history of the Batoidea (Fig. 3).

METHODOLOGYSpecimen Collection

Specimens of nondurophagous taxa, Raja eglan-teria (n 5 2), Rhinobatos lentiginosus (n 5 1),

Gymnura micrura (n 5 3), Dasyatis sabina (n 5 3),Urobatis jamaicensis (n 5 2) and durophagoustaxa, Rhinoptera bonasus (n 5 5), Aetobatus nari-nari (n 5 2), and Myliobatis freminvillei (n 5 5),were collected primarily via fisheries-independentsurveys in collaboration with several agencies;Florida Fish and Wildlife Conservation Commis-sion (Charlotte Harbor and Eastpoint, FLA),National Marine Fisheries Service GulfSPAN sur-veys, and through the Virginia Institute ofMarine Science (Gloucester Point, VA). Specimenswere typically frozen. Animals considered for thisstudy showed no signs of undue stress or unusualcharacteristics of the cranial region. For simplic-ity, only the genus will be referenced, as only onespecies representing each genus was usedthroughout the study. Animals were euthanizedby severing of the vertebral column near thechondrocranium, overdosing of MS-222, or byplacing the animal on ice in accordance with Ani-mal Care and Use Committee guidelines at Flor-ida State University (Protocol 1209). Referencedesignations for specimens used to inform illus-trations are as follows: Rajidae: Raja eglanteria,FSU/REGA—002, DW 5 37.5 cm (male); Rhinoba-tidae: Rhinobatos lentiginosus, FSU/RLEN—001,DW 5 16.5 cm, total length 5 56 cm (female);Gymnuridae: Gymnura micrura, FSU/GMIC—002, DW 5 58.5 cm (female); Rhinopteridae: Rhi-noptera bonasus, FSU/RBON—043, DW 5 78.2 cm(female); Myliobatidae: Aetobatus narinari, FSU/ANAR—001, DW 5 99 cm (female); Myliobatidae:Myliobatis freminvillei, FSU/MFRE—002,DW 5 72 cm (male); Dasyatidae: Dasyatis sabina,FSU/DSAB—002, DW 5 28.5 cm (female); andUrolophidae: Urobatis jamaicensis, FSU/UJAM—001, DW 5 21 cm (female). Additional specimeninformation available upon request.

Terminology

Anatomical data were organized according toembryonic muscle units, as designated by Edge-worth (1935), Miyake et al. (1992), Lovejoy (1996),and Mallat (1997); data from previous works weresynthesized (with some reinterpretation asneeded) to provide context and broader perspectivefor our data (see Tables 1 and 2). Muscles wereidentified with consideration of both attachment(origin and insertion) as well as general position.Raja eglanteria, Rhinobatos lentiginosus, Gym-nura micrura, Dasyatis sabina, and Urobatisjamaicensis are referred to below as“nondurophagous” taxa, whereas Rhinoptera bona-sus, Aetobatus narinari, and Myliobatis freminvil-lei are “durophagous.” Also, the term “stingray”includes all species examined except Raja eglante-ria and Rhinobatos lentiginosus (see supplemen-tary Information for details about batoid cranialmuscle terminology).

Fig. 3. Phylogenetic relationships of surveyed taxa. Modifiedfrom Aschliman et al. (2012). Branch lengths not to scale.Speciesused in this study in italics.



RESULTSSkeletal Overview

The range of variation in chondrichthyan cranialskeletal morphology is broad and covered in greatdetail in other works (e.g., see Didier, 1995; Love-joy, 1996; Nishida, 1990, and particularly Claeson,2011 for more thorough treatment). Here, we pro-vide a generalized summary of the batoid cranialskeleton to illustrate the structural framework on

which the muscles attach; a visual comparison ofchondrichthyan cranial skeletons is provided inFigure 1, with generalized muscles mapped on tothe batoid model in Figure 2, and with a moredetailed generalized batoid skeleton illustrated inFigure 4.

Batoid fishes are typically flatter than sharksand, therefore, the basic layout of the cranial skel-eton can be largely understood in a two-

TABLE 1. Cranial musculature in eight species of batoids

Abbreviation Division Found in

Generalized attachmentGeneralfunctionOrigin Insertion

AMMe Adductor mandibulaemedialis

Re, Gm, An,Mf, Rb, Ds, Uj

Palatoquadrate Meckel’s cartilage Jaw closure

AMLa Adductor mandibulaelateralis

all Palatoquadrate Meckel’s cartilage Jaw closure

AMMa Adductor mandibulaemajor

all Palatoquadrate Meckel’s cartilage Jaw closure

AMD Adductor mandibulaedeep

An, Mf, Rb Palatoquadrate Meckel’s cartilage Jaw closure

AMLi Adductor mandibulaelingualis

Gm, Rb Palatoquadrate Meckel’s cartilage Jaw closure

SP Spiracularis all Postorbital—chondrocranium

Spiracular cartilage,palatoquadrate

Ventilation

LP Levator palatoquadrati all Preorbital—chondrocranium

Palatoquadrate Jaw retraction

SB Suborbitalis all Palatoquadrate Meckel’s cartilage Protrudes jawsPCM Precranial muscle Gm, An, Mf Chondrocranium Pectoral propterygium UnknownETM Ethmoideo-parethmoidalis Gm, An, Mf,

Rb, Ds, UjChondrocranium Pectoral propterygium Unknown

CHV Constrictor hyoideusventralis

all Medial region—1st gill arch

Ventral 1st gill arch Ventilation

CHD Constrictor hyoideusdorsalis

all Medial region—1st gill arch

Dorsal 1st gill arch Ventilation

LHYM Levator hyomandibularis all Dorsal chondrocranium Hyomandibularcartilage

Retracts hyoidcartilages

DHYM Depressor hyomandibularis all Depressor rostri Hyomandibularcartilage

Depresseshyoid arch

DM Depressor mandibularis Ds, Uj Depressor rostri Meckel’s cartilage Depresseslower jaw

DR Depressor rostri all Antimere—ventralmidline

Pectoral propterygium Lowers snout

LR Levator rostri Re, Rl Dorsal chondrocranium Pectoral propterygium Raises snoutCM Coracomandibularis all Coracoarcualis Meckel’s cartilage Abducts jawsCH Coracohyoideus all Coracoarcualis Basihyal, hypohyal

cartilagesExpands

oropharynxCARC Coracoarcualis all Pectoral girdle Coracomandibularis Abducts jaws

(w/ CM)CHYM Coracohyomandibularis all Antimere—ventral

midlineHyomandibular

cartilagesDepresses

hyoid arch

Abbreviations: Re, Raja eglanteria; Rl, Rhinobatos lentiginosus; Gm, Gymnura micrura; An, Aetobatus narinari; Mf, Myliobatisfreminvillei; Rb, Rhinoptera bonasus; Ds, Dasyatis sabina; Uj, Urobatis jamaicensis.

TABLE 2. Generalized chondrichthyan cranial muscle function

Abbreviation Muscle unit General function

Number of muscle divisions

Ratfishes Sharks Batoids

AM Adductor mandibulae Jaw closure 3 2–5 3–6CD Constrictor dorsalis Jaw suspension 1 1–2 2CH Constrictor hyoideus Hyoid suspension 2 3 5–6GH Geniohyoideus Abducts jaws 1 1 1RC Rectus cervicus Abducts jaws 1 2 3



dimensional, ventral view. There are typically 10skeletal elements (counting paired cartilages astwo independent elements—Figs. 1 and 4). Thechondrocranium can exhibit complex, convolutededges, but is essentially a rectangular braincase,with bulbous anterior nasal capsules and in somecases (e.g., skates, guitarfishes) a long rostral car-tilage at the anterior end of the animal. The neu-rocranial fontanelle is a slight depression alongthe dorsal medial axis of the chondrocranium,within which a very thin cartilaginous sheath cov-ers access to the brain cavity. The long, slenderhyomandibular cartilages articulate with the lat-eral otic regions near the caudal end of the brain-case, extending anterolaterally toward thepropterygia (see below) and articulating via stoutligaments with the extreme lateral edges of thejaw complex. The spiracle, a small orifice on thedorsal body surface, opening via a short passage-way into the pharynx, is located in the triangularspace between the anterior edge of the hyomandi-bulae, the caudal edges of the jaws and the lateraledge of the cranium. The shaft of the hyomandib-ula and a thin, bowed spiracular cartilage formthe posterior and anterior margins of the spiracle,respectively. The spiracle represents the reducedgill-slit opening between the mandibular andhyoid arches.

Although the batoids exhibit a range of dietaryspecializations, in all cases, the upper jaw (PQ)and lower jaw (MK) can be considered to simplyspan the space between the two distal hyomandib-ular tips, with all nondurophagous taxa (exceptMobula and Manta) having unfused jaw symphy-ses. The teeth, whether cuspidate and arranged onpads (as in nondurophagous stingrays) or flat andpavement like (as in pelagic/durophagous sting-

rays), overlie the jaw symphyses. The wing pro-cess, a ventral projection of the MK, is found onlyin durophagous stingrays (Summers, 2000).

The caudal end of the braincase articulates withthe synarcual, a long tube-like element believed tobe derived from fused elements of the first severalvertebrae (Claeson, 2011) and joins the vertebralcolumn caudally. This region is flanked by the“branchial basket,” composed of a series of rostro-caudal jointed skeletal arches—composed of thesingle pseudohyoid and multiple branchialarches—extending like medially oriented crescentsfrom the posterior braincase and synarcual tomeet on the ventral pharyngeal floor in a singlefused basibranchial copula. In this way, the bran-chial basket superficially resembles a ribcage sur-rounding the pharynx, with the gill pouches andgill slits between each consecutive arch. In nonpe-lagic stingrays, the basibranchial copula also artic-ulates anteriorly via hypohyal cartilages with thesole additional portion of the hyoid arch, thebasihyal cartilage; this ventral hyoid element isabsent in pelagic stingrays.

In addition to forming the roof of the branchialbasket, the synarcual articulates caudally with themedial portion of the pectoral girdle (including thescapulocoracoid and coracoid cartilage) in nontor-pediniform batoids. This portion of the pectoralgirdle spans lateromedially, perpendicular to thesynarcual and the animal’s long axis. The pectoralpropterygia, long scythe-shaped cartilages, archanteriorly from their articulation with the lateralpectoral girdle to meet with the antorbital carti-lages or rostral cartilages, effectively “framing” thecranial elements rostral of the coracoid bar. Indurophagous stingrays, the dorsal surfaces of thepropterygia are fused to the ventral chondrocra-nium. Fin radials, small skeletal elements sup-porting the pectoral fins, extend laterally from thepropterygia (Mulvany and Motta, 2013). For ourpurposes, the propterygia are considered to formthe lateral boundaries of the “head” of the batoidsin this study, including the chondrocranium–syn-arcual complex as well as the associated jaw,hyoid, and branchial arches.

Hypaxial Series

This muscle series develops embryonically fromthe genio-hyoideus and rectus cervicus musclegroups (Tables 1 and 2, Miyake et al., 1992). Threeof these muscles, the coracoarcualis (CARC or rec-tus cervicus, Nishida, 1990), coracomandibularis(CM or genio-coracoideus, Nishida, 1990), and cor-acohyoideus (CH), are present in all chon-drichthyans, with the coracohyomandibularis(CHYM) being unique to batoids (Miyake et al.,1992; Anderson, 2008). Although muscle configura-tions differ considerably across vertebrate groupsin terms of their attachment, the hypaxial muscles

Fig. 4. Generalized anterior batoid skeletal structure.



generally originate on some aspect of the pectoralgirdle (or another hypaxial muscle, which theninserts on the girdle) and insert anteriorly on thelower jaw or hyoid apparatus (Edgeworth, 1935;Miyake et al., 1992; Anderson, 2008). The hypaxialmuscles are associated ancestrally in gnathos-tomes with jaw-opening (lower jaw abduction) andpull the feeding apparatus posteriorly and ven-trally (Mallat, 1997; Anderson, 2008).

CARC and CM. In all taxa, the CARC (Figs.6–8, all taxa) are paired, triangular muscles whichtypically originate on the antero-ventral surface ofthe coracoid bar, superficial to the ventral surfaceof the pericardium and insert on the CM. Paired,anterior CM antimeres, aligned in series with theposterior CARC muscles (Figs. 6 and 7 and 9 and10) together abduct the lower jaw (Huber et al.,2011). In batoids overall, the arrangement, attach-ment, and presumably the function of the CM andCARC are generally conserved (Raja, Figs. 6 and7a and 9 and 10a; Rhinobatos, Figs. 6 and 7b and9 and 10b; Gymnura, Figs. 6 and 7c and 9 and10c; Dasyatis, Figs. 6 and 7g and 9 and 10g; andUrobatis, Figs. 6 and 7h and 9 and 10h). TheCARC muscle is reduced in Myliobatis (Figs. 6–8e). Our specimens of Aetobatus had been severedanterior to the origin of the CARC, precluding fullexamination of this muscle. Gonzalez-Isais (2003)described the condition of the CARC as being gen-erally similar in Aetobatus to the condition wehave described in Myliobatis and Rhinoptera(Figs. 6–8f).

CH. The CH in most taxa typically originateson the antero-medial face of the first gill arch andthe CARC and inserts on the underside of the pha-ryngeal region and ventral aspects of the basihyaland hypohyal cartilages (Figs. 7 and 8). The CH isan embryonic derivative of the rectus cervicus(Miyake et al., 1992) and depresses the hypohyaland basihyal cartilages, which in turn expand thepharyngeal cavity (Huber et al., 2011). The CH isnoticeably small and more strap like in Rhinobatos(Fig. 8b) than in stingrays, articulating less withthe first gill arch and more with the fascia formingthe junction of the CHYM and CARC. In Gymnura(Fig. 7c), Dasyatis (Fig. 7g), and Urobatis (Fig. 7h)the CH is also found dorsal to the CM, flanking iton both sides and originating on the combinedbasihyal/hypohyal cartilages. The CH is reportedto be absent in pelagic myliobatiforms (Miyakeet al., 1992). However, other authors (Nishida,1990; Lovejoy, 1996) consider the “Y” muscle ofMiyake et al. (1992) to be homologous to the CH.This muscle (CH in our diagrams) is found in Rhi-noptera (Fig. 8f), Aetobatus (Figs. 7 and 8d), andMyliobatis (Figs. 7 and 8e), on the ventral aspectof the extrabranchial cartilages (in a manner con-sistent with the CH in other taxa—Miyake et al.,1992), stretching transversely across the first gill

arches and meeting medio-ventrally with theCHYM.

CHYM. The CHYM is a paired ventral muscleunique to batoids (Miyake et al., 1992), usuallysuperficial to the lateral portions of the ventralpharyngeal region. In stingrays, the CHYM arecolumnar at their origins on the pericardium,CARC, and coracoid, extending along the ventralpharyngeal cavity and inserting on the hyoman-dibular cartilages. The function of the CHYM pre-sumably involves depression and abduction of thehyomandibular cartilages (and to some extent, thelower jaw and pharyngeal region as well) ventrallyand medially (Dean and Motta, 2004a). TheCHYM is similar to the depressor hyomandibula-ris (DHYM) in depressing the hyoid cartilages ven-trally, but differs in its origin being placed moreposteriorly, therefore, likely retracts the hyoman-dibulae posteriorly as well as ventrally. In Raja(Fig. 8a) and Rhinobatos (Fig. 8b) the CHYM origi-nates directly on the hyomandibular cartilages,without a tendon. In stingrays the CHYM insertson the base of the tissue of the pharynx and thebasihyal cartilages, inserting on the distal aspectof the ventral hyomandibular cartilages via longtendons (Figs. 7 and 8c). In Rhinoptera (Fig. 8f),Aetobatus (Fig. 8d), Myliobatis (Fig. 8e), Dasyatis(Figs. 7 and 8g), and Urobatis (Figs. 7 and 8h), theCHYM extend laterally over much of the ventralpharynx region, going so far Aetobatus, Rhinoptera(Fig. 8) and Myliobatis, to as line the dorsal sur-face of the MK. The arrangement of the CHYM inAetobatus is slightly different, with the muscleconsisting of two divisions which both insert apo-neurotically on the jaw joint ligament and theentire dorsal surface of the MK, respectively.

Hyoid Series

Mallat (1997) postulates that the embryonic orearly evolutionary function of the branchial con-strictor muscles was primarily expiration. Withthe evolution of jaws in gnathostomes, the musclesof the first two gill arches (mandibular and hyoid,respectively) changed their function correspond-ingly. Three muscles stemming from the hyoidmusculature are found in all elasmobranchs (Fig.5), the levator hyomandibularis (LHYM), constric-tor hyoideus ventralis (CHV) and constrictor hyoi-deus dorsalis (CHD—this muscle, however, isabsent, perhaps secondarily, in holocephalans—Miyake et al., 1992). The depressor rostri (DR),DHYM, and levator rostri (LR) are unique tobatoids with the latter found in skates, guitar-fishes, and torpediniforms and presumably lost inmyliobatiform stingrays (Miyake et al., 1992; Deanand Motta, 2004a). The depressor mandibularis(DM) is only found in dasyatoid rays (Dasyatis,Urobatis). The LHYM elevates the hyomandibularcartilages in all taxa (Mallat, 1997), specifically, in



those species in which it has been studied experi-mentally, lifting the hyomandibular cartilages dur-ing the recovery phase of jaw protrusion towardthe chondrocranium in taxa (Wilga and Motta,1998). The CHV and CHD compress the first gillpouches (Huber et al., 2011). The DR and LR areantagonists and depress and elevate the snout,respectively. The DR also presumably serves tocompress the cephalic and branchial region duringthe preparatory phase of prey capture (Wilga andMotta, 1998). The DHYM adduct the hyomandibu-lar cartilages ventrally during feeding, are associ-ated with subsequent jaw protrusion, andantagonize LHYM (Wilga and Motta, 1998). TheDM aids in retracting the lower jaw cartilagesduring feeding (Miyake et al., 1992).

LHYM. The LHYM originates on the lateralaspect of the dorsal surface of the chondrocra-nium, just posterior to (although sometimes over-lapping) the postorbital cartilages (Fig. 5, all taxa)

and inserting on the dorsal and posterolateralsurfaces of the hyomandibular cartilage. This mus-cle abuts the CHD through shared connective tis-sue. The condition and presumed function of theLHYM appears broadly conserved across the taxasurveyed in this study. In Aetobatus, the LHYMmuscles are covered by an overlying layer of tissueand are sunken within fossae formed by theposterior-lateral portion of the otic region andbounded laterally by the first gill arch (Fig. 5e).

LR. The LR originates on the dorsal surface ofthe cervicothoracic synarcual, lateral to the epax-ial muscles, and superficial to the LHYM. The LRinserts on the dorsal surface of the pectoral prop-terygium. The LR is lacking in all myliobatiformrays surveyed, but is present in Raja (Fig. 5a) andRhinobatos (Fig. 5b). The LR inserts on the dorsalsurface of the pectoral propterygium, a configura-tion reminiscent of the DR insertion on the ventralsurface of the propterygium. The presumed loss of

Fig. 5. Dorsal superficial cephalic musculature of six species of batoid fishes. All crania with anterior facing left. (a) Raja eglante-ria, (b) Rhinobatos lentiginosus, (c) Gymnura micrura, (d) Aetobatus narinari, (e) Myliobatis freminvillei, (f) Rhinoptera bonasus, (g)Dasyatis sabina, (h) Urobatis jamaicensis; AMD, Adductor mandibulae deep; AMMa, Adductor mandibulae major; AMMe, Adductormandibulae medialis; AMLa, Adductor mandibulae lateralis; AMLi, Adductor mandibulae lingualis; CARC, Coracoarcualis; CH, Cora-cohyoideus; CHD, Constrictor hyoideus dorsalis; CHV, Constrictor hyoideus ventralis; CHYM, Coracohyomandibularis; CM, Coraco-mandibularis; DHYM, Depressor hyomandibularis; DM, Depressor mandibularis; DR, Depressor rostri; ETM, Ethmoideo-parethmoidalis; LHYM, Levator hyomandibularis; LP, Levator palatoquadrati; LR, Levator rostri; PQ, Palatoquadrate; PCM, Precra-nial muscles; MK, Meckel’s cartilage; SB, Suborbitalis; SP, Spiracularis. Muscles in red are derived from the adductor mandibulae(mandibular plate); muscles in green are derived from the constrictor dorsalis (mandibular plate); muscles in yellow are derived fromthe constrictor hyoideus (hyoid plate); muscles in orange are derived from the hypaxial plate (genio-hyoideus and rectus cervicus);the SB is shaded purple to highlight its unique position (oral/premandibular plate), other oral/premandibular muscles are shaded inblue.



the LR in stingrays may be correlated with thereduction of the rostral cartilages and snout rigid-ity in these taxa in general (as opposed to skatesand guitarfish).

CHV/CHD. The CHV originates on the mid-ventral hypobranchial septa (Fig. 6), whereas theCHD originates on the dorsal aspect of the firstgill arch. In all species surveyed the CHD insertsalong the fascia it shares with its ventral compo-nent (Fig. 5). In this manner, the CHV and CHVcover the most anterior transverse surface of thefirst gill arches, forming a muscular wall separat-ing the lateral hyoid and branchial regions.

DR. The DR generally originates on a midven-tral raphe, superficial to the hypaxial musculature

and inserts on the rostrum, either directly orthrough a broad tendinous sheath of connectivetissue. In some species, the DR covers nearly theentire ventral area between the anterior branchialbasket and the lower jaw. The DR in Raja (Figs. 6and 7a) and Rhinobatos (Fig. 6b), are conspicu-ously narrow, originating on the fascia ventral tothe CARC and inserting via a long, thin tendon tothe pectoral propterygium. In Gymnura, the DRare bilateral muscles which originate on fasciaassociated with the anterior portion of the CARCmuscles and insert on the superficial fascia cover-ing the ventral surface of the anterior portion ofthe pectoral propterygium (Fig. 6c). In Dasyatis(Fig. 6g) and Urobatis (Fig. 6h) the DR are thin

Fig. 6. Ventral superficial cephalic musculature of six species of batoid fishes; anterior is to the top. (a) Raja eglanteria, (b) Rhino-batos lentiginosus, (c) Gymnura micrura, (d) Aetobatus narinari, (e) Myliobatis freminvillei, (f) Rhinoptera bonasus, (g) Dasyatissabina, (h) Urobatis jamaicensis; AMD, Adductor mandibulae deep; AMMa, Adductor mandibulae major; AMMe, Adductor mandibu-lae medialis; AMLa, Adductor mandibulae lateralis; AMLi, Adductor mandibulae lingualis; CARC, Coracoarcualis; CH, Coracohyoi-deus; CHD, Constrictor hyoideus dorsalis; CHV, Constrictor hyoideus ventralis; CHYM, Coracohyomandibularis; CM,Coracomandibularis; DHYM, Depressor hyomandibularis; DM, Depressor mandibularis; DR, Depressor rostri; ETM, Ethmoideo-parethmoidalis; LHYM, Levator hyomandibularis; LP, Levator palatoquadrati; LR, Levator rostri; PQ, Palatoquadrate; PCM, Precra-nial muscles; MK, Meckel’s cartilage; SB, Suborbitalis; SP, Spiracularis. Muscles in red are derived from the adductor mandibulae(mandibular plate); muscles in green are derived from the constrictor dorsalis (mandibular plate); muscles in yellow are derived fromthe constrictor hyoideus (hyoid plate); muscles in orange are derived from the hypaxial plate (genio-hyoideus and rectus cervicus);the SB is shaded purple to highlight its unique position (oral/premandibular plate), other oral/premandibular muscles are shaded inblue.



and insert on the propterygium via connective tis-sue. The DR insert on the rostral region of the cra-nium or the rostral snout/cephalic lobe muscles indurophagous taxa and presumably serves to com-press the cephalic and branchial regions duringforaging (Sasko et al., 2006; Mulvany and Motta,2013).

DHYM. The DHYM originates on a shared,midline raphe attached to the dorsal surface of theCM and extends laterally to insert on the postero-lateral aspect of the hyomandibular cartilages.The antimeres of the DHYM in skates (Raja, Fig.7a) do not share a common medial origin, insteadoriginating lateral and slightly dorsal to eitherside of the CM. In guitarfish (Rhinobatos, Fig. 7b),

the DHYM divisions originate on the ventral mid-line, attaching to the dorsal side of the CM. InGymnura (Fig. 7c), the two triangular-shaped anti-meres of the DHYM share a common raphe at theventral body midline, ventral to the CM and dor-sal to the DR and inserting on the hyomandibularcartilage via tendons. In Dasyatis (Figs. 6 and 7g)and Urobatis (Figs. 6h), the DHYM inserts directly(without a tendon) on the hyomandibular carti-lages. In Rhinoptera (Fig. 8f), Aetobatus (Fig. 8d),and Myliobatis (Figs. 7 and 8e), the DHYM origi-nates on the medial, postero-dorsal side of the DRand inserts on the postero-lateral surface of thehyomandibular cartilages via a long tendon. TheDHYM in Myliobatis and Aetobatus is narrower

Fig. 7. Ventral middeep cephalic musculature of six species of batoid fishes; anterior is to the top. (a) Raja eglanteria, (b) Rhinoba-tos lentiginosus, (c) Gymnura micrura, (d) Aetobatus narinari, (e) Myliobatis freminvillei, (f) Rhinoptera bonasus, (g) Dasyatissabina, (h) Urobatis jamaicensis; AMD, Adductor mandibulae deep; AMMa, Adductor mandibulae major; AMMe, Adductor mandibu-lae medialis; AMLa, Adductor mandibulae lateralis; AMLi, Adductor mandibulae lingualis; CARC, Coracoarcualis; CH, Coracohyoi-deus; CHD, Constrictor hyoideus dorsalis; CHV, Constrictor hyoideus ventralis; CHYM, Coracohyomandibularis; CM,Coracomandibularis; DHYM, Depressor hyomandibularis; DM, Depressor mandibularis; DR, Depressor rostri; ETM, Ethmoideo-parethmoidalis; LHYM, Levator hyomandibularis; LP, Levator palatoquadrati; LR, Levator rostri; PQ, Palatoquadrate; PCM, Precra-nial muscles; MK, Meckel’s cartilage; SB, Suborbitalis; SP, Spiracularis. Muscles in red are derived from the adductor mandibulae(mandibular plate); muscles in green are derived from the constrictor dorsalis (mandibular plate); muscles in yellow are derived fromthe constrictor hyoideus (hyoid plate); muscles in orange are derived from the hypaxial plate (genio-hyoideus and rectus cervicus);the SB is shaded purple to highlight its unique position (oral/premandibular plate), other oral/premandibular muscles are shaded inblue.



than in the nondurophagous species and pennatefibered in Rhinoptera (Fig. 8). The DHYM in Aeto-batus is considerably reduced compared to Rhinop-tera and Myliobatis. In Rhinoptera, a division ofthe DHYM runs along the postero-lateral surfaceof the hyomandibular cartilage, overlying theCHYM.

DM. In those taxa where the DM was present,the muscle originates on the anterior edge of theDHYM (just dorsal to the DR, in most cases) andinserts directly on the most postero-lateral processof the MK, just posterior to the hyomandibulararticulation with the jaws. The DM is associatedwith the DHYM and may not represent a distinctmuscle, and its presence is variable in batoids(Miyake et al., 1992). The DM was found conclu-

sively in Dasyatis (Figs. 9 and 10g) and Urobatis(Figs. 9 and 10h), but was not located in Raja,Rhinobatos or Gymnura (where it is expected—Miyake et al., 1992). The DM is lacking altogetherin myliobatids and rhinopterids (Miyake et al.,1992).

Oral/Precranial Series

These muscles are the vestige (along with thelabial cartilages and lip folds) of the pregnathos-tome “oral” mouth (Mallat, 1996, 1997), the jawsof modern gnathostomes representing a pharyn-geal or branchial-derived mouth. The specific rela-tionship between the ethmoideo-parethmoidalis(ETM), precranial muscles (PCM), and suborbitalis[SB 5 levator labii superioris, Edgeworth (1935)

Fig. 8. Ventral deep cephalic musculature of six species of batoid fishes; anterior is to the top. (a) Raja eglanteria, (b) Rhinobatoslentiginosus, (c) Gymnura micrura, (d) Aetobatus narinari, (e) Myliobatis freminvillei, (f) Rhinoptera bonasus, (g) Dasyatis sabina,and (h) Urobatis jamaicensis; AMD, Adductor mandibulae deep; AMMa, Adductor mandibulae major; AMMe, Adductor mandibulaemedialis; AMLa, Adductor mandibulae lateralis; AMLi, Adductor mandibulae lingualis; CARC, Coracoarcualis; CH, Coracohyoideus;CHD, Constrictor hyoideus dorsalis; CHV, Constrictor hyoideus ventralis; CHYM, Coracohyomandibularis; CM, Coracomandibularis;DHYM, Depressor hyomandibularis; DM, Depressor mandibularis; DR, Depressor rostri; ETM, Ethmoideo-parethmoidalis; LHYM,Levator hyomandibularis; LP, Levator palatoquadrati; LR, Levator rostri; PQ, Palatoquadrate; PCM, Precranial muscles; MK, Meck-el’s cartilage; SB, Suborbitalis; SP, Spiracularis. Muscles in red are derived from the adductor mandibulae (mandibular plate);muscles in green are derived from the constrictor dorsalis (mandibular plate); muscles in yellow are derived from the constrictorhyoideus (hyoid plate); muscles in orange are derived from the hypaxial plate (genio-hyoideus and rectus cervicus); the SB is shadedpurple to highlight its unique position (oral/premandibular plate), other oral/premandibular muscles are shaded in blue.



or 5 preorbitalis, Mallat (1996, 1997)] is unknownand more generally, the embryonic origin of thetwo former muscles is entirely uncertain (we sug-gest here, based on placement and orientationthat they may be precranial). The oral/PCM,including the SB, precranial, and ETM, likely dealwith retraction and possibly controlled realign-ment of the jaw-adductor muscle complex backtoward the cranium during jaw protrusion (Mallat,1997), or more generally with maintaining ante-rior contact between the cranium and jaws. Boththe ETM and PCM are miniscule in Raja and Rhi-nobatos, but seem to articulate the jaw adductors(and the jaws themselves) to the pectoral proptery-gium. It is possible that the PCM and ETMmuscles may simply be different divisions of thesame muscle, however, it is possible in most spe-

cies to differentiate these two divisions based onfiber direction.

ETM. The ETM muscles in Gymnura originateon the antero-lateral regions of the chondrocra-nium, just ventral to the eye (see Appendix formore information). They insert on the disto-lateralpectoral propterygium cartilage, which encirclesthe cranial region in these stingrays, runningalong the medio-lateral surface, intimately associ-ating with the PCM (Fig. 8). In Myliobatis, thePCM and ETM muscles are difficult to reveal, asthey are found on the medial surface of the pecto-ral propterygium (Gonzalez-Isais, 2003). The ETMin this species originates on the medio-lateral sur-face of the posterior PCM and inserts onto the jawadductor musculature, in particular the adductormandibulae deep’s (AMD) point of origin on the

Fig. 9. Dorsal jaw adducting musculature of six species of batoid fishes; anterior is to the top. (a) Raja eglanteria, (b) Rhinobatoslentiginosus, (c) Gymnura micrura, (d) Aetobatus narinari, (e) Myliobatis freminvillei, (f) Rhinoptera bonasus, (g) Dasyatis sabina,and (h) Urobatis jamaicensis; AMD, Adductor mandibulae deep; AMMa, Adductor mandibulae major; AMMe, Adductor mandibulaemedialis; AMLa, Adductor mandibulae lateralis; AMLi, Adductor mandibulae lingualis; CARC, Coracoarcualis; CH, Coracohyoideus;CHD, Constrictor hyoideus dorsalis; CHV, Constrictor hyoideus ventralis; CHYM, Coracohyomandibularis; CM, Coracomandibularis;DHYM, Depressor hyomandibularis; DM, Depressor mandibularis; DR, Depressor rostri; ETM, Ethmoideo-parethmoidalis; LHYM,Levator hyomandibularis; LP, Levator palatoquadrati; LR, Levator rostri; PQ, Palatoquadrate; PCM, Precranial muscles; MK, Meck-el’s cartilage; SB, Suborbitalis; SP, Spiracularis. Muscles in red are derived from the adductor mandibulae (mandibular plate);muscles in green are derived from the constrictor dorsalis (mandibular plate); muscles in yellow are derived from the constrictorhyoideus (hyoid plate); muscles in orange are derived from the hypaxial plate (genio-hyoideus and rectus cervicus); the SB is shadedpurple to highlight its unique position (oral/premandibular plate), other oral/premandibular muscles are shaded in blue.



PQ (Fig. 8e). The PCM muscles in Aetobatus areeither reduced or entirely absent, the ETM arethin (Fig. 8d). The ETM in Rhinoptera are foundmedial and dorsal to the origin of the cephalic lobemusculature, ventral to the inferior eye muscles.The ETM in Rhinoptera originates on the ventralside of the rostral cartilage, anterior and slightlydorsal to the cartilage surrounding the nares (Fig.8f). The ETM inserts on the pectoral propterygiumjust ventral to the posterior-most region of the spi-racular opening (Figs. 4 and 6–8). In Dasyatis(Figs. 7 and8g) and Urobatis (Figs. 7 and 8h), onlythe ETM is readily distinguishable, associatingclosely with the jaw adductor musculature and theanterior pectoral propterygium. In both Dasyatisand Urobatis the ETM originates on the lateralsurface of the anterior neurocranial region and

inserts directly on the jaw adductor musculatureand the pectoral propterygium.

PCM. The left and right PCM divisions inGymnura (Figs. 7 and 8c) originate on the rostralmidline at the anterior surface of the chondrocra-nium (anterior to the fontanelle) and proceed lat-erally, hugging the postero-medial surface of theanterior pectoral propterygium, eventually insert-ing on the lateral corner of the chondrocranium,ventral to the preorbital processes (Nishida, 1990;Gonzalez-Isais, 2003). The PCM in Myliobatisunderlie the inferior muscles of the eye (and partlythe eye itself), inserting on the posterior-ventralsurface of the rostral cartilages. For Aetobatus, seeETM section above. The PCM in Rhinoptera areeither reduced or indistinguishable from the ETMmuscles. Further inquiry into the function,

Fig. 10. Ventral jaw adducting musculature of six species of batoid fishes; anterior is to the top. (a) Raja eglanteria, (b) Rhinobatoslentiginosus, (c) Gymnura micrura, (d) Aetobatus narinari, (e) Myliobatis freminvillei, (f) Rhinoptera bonasus, (g) Dasyatis sabina,and (h) Urobatis jamaicensis; AMD, Adductor mandibulae deep; AMMa, Adductor mandibulae major; AMMe, Adductor mandibulaemedialis; AMLa, Adductor mandibulae lateralis; AMLi, Adductor mandibulae lingualis; CARC, Coracoarcualis; CH, Coracohyoideus;CHD, Constrictor hyoideus dorsalis; CHV, Constrictor hyoideus ventralis; CHYM, Coracohyomandibularis; CM, Coracomandibularis;DHYM, Depressor hyomandibularis; DM, Depressor mandibularis; DR, Depressor rostri; ETM, Ethmoideo-parethmoidalis; LHYM,Levator hyomandibularis; LP, Levator palatoquadrati; LR, Levator rostri; PQ, Palatoquadrate; PCM, Precranial muscles; MK, Meck-el’s cartilage; SB, Suborbitalis; SP, Spiracularis. Muscles in red are derived from the adductor mandibulae (mandibular plate);muscles in green are derived from the constrictor dorsalis (mandibular plate); muscles in yellow are derived from the constrictorhyoideus (hyoid plate); muscles in orange are derived from the hypaxial plate (genio-hyoideus and rectus cervicus); the SB is shadedpurple to highlight its unique position (oral/premandibular plate), other oral/premandibular muscles are shaded in blue.



relationship, and embryological derivation of thePCM and ETM are needed, particularly consider-ing their variable presence and morphology amongspecies and nonintuitive function.

SB. The SB originates on the ventral side ofthe chondrocranium and inserts on the MK via anarrow tendon, likely functioning in Raja (Fig.9a), Rhinobatos (Figs. 9 and 10b), Gymnura (Figs.8c, 9, and 10c), Dasyatis (Fig. 10g), and Urobatis(Fig. 10h) as a means of lifting the jaws towardthe chondrocranium. In Rhinoptera (Fig. 10f),Aetobatus (Fig. 10d), and Myliobatis (Figs. 9 and10e), the SB has become disassociated from thechondrocranium and likely functions instead injaw adduction. The SB in these durophagous taxais a relatively large, parallel-fibered, delta-shapedmuscle originating on the latero-ventral surface ofthe PQ, inserting via a tendon [shared with theadductor mandibulae lateralis (AMLa) complex] onthe anterior-most lateral portion of the wing pro-cess on the MK. Consistent insertion of this mus-cle via a tendon on the lower jaw is commonacross all taxa as well as its innervation by themandibular nerve V3 (Mallat, 1996, 1997).

Mandibular Series

The muscles of the mandibular series arederived from either the embryonic constrictor dor-salis or adductor mandibulae. These musclesevolved from gill constrictors on pregnathostomeanterior gill arches (but see Supporting Informa-tion). In elasmobranchs, the constrictor dorsalisgives rise to two muscles: the levator palatoqua-drati (LP) and the spiracularis (SP). In batoids,the embryonic adductor mandibulae gives rise to alarge number of muscle divisions, namely theadductor mandibulae medialis (AMMe), AMLa,AMD, adductor mandibulae lingualis (AMLi) andthe adductor mandibulae major (AMMa). Theintermandibularis of sharks is not present inBatoidea. Muscles of the constrictor dorsalis (LPand SP) originate on the ventral surface of thechondrocranium and insert on the PQ (to someextent). Specifically, the SP originates posterior tothe LP and inserts along the rostral portion of thespiracular cartilage as well as portions of the ante-rior and dorsal PQ. The muscles of the AM com-plex all originate on the ventral PQ cartilage andinsert on the ventral MK (but see AMLi). The LPand SP aid in retraction of the PQ toward thechondrocranium. The SP specifically rotates thespiracular cartilage to expose the spiracular open-ing. The adductor mandibulae muscles predomi-nantly adduct the upper and lower jaws.

LP. The LP is always closely aligned with theSP and distinguishing where one muscle beginsand the other ends is sometimes difficult. In allspecies surveyed the LP originates on the ventralface of the chondrocranium, on the medial surface

of the otic region, posterior and perpendicular(transverse) to the origin of the SB in specieswhere the SB still maintains contact with the cra-nium (see above), but still anterior to the origin ofthe SP. In all taxa surveyed, the LP inserts on thesoft tissue surrounding the lingual interior of thejaws, dorsal to the PQ as well as directly attachingto the PQ itself (Figs. 9 and10). The LP is notice-ably smaller reduced in Myliobatis than in Rhi-noptera. In Rhinoptera and Aetobatus the LPconsists of three muscle heads, two of which origi-nate medially within the otic region (one slightlyanterior to the other) and the third originating farposterior to the other two and lining the dorsal tis-sue of the pharyngeal chamber (Figs. 9 and 10 d–f). In Aetobatus and Myliobatis the LP insertsdirectly on the dorsal anterior edge of the PQ witha tendon and aponeurotically to the surroundingoral tissue (Fig. 9e).

SP. The SP is generally a small, strap-likemuscle that originates on the lateral aspect ofchondrocranium, runs adjacent (medio-posteriorly)to the LP, inserts along the dorsal surface of thespiracular cartilage and wraps around the carti-lage until it covers both the posterior and anteriorfaces. In all batoids except the durophagous taxa(Fig. 5), the SP predominantly covers the anteriorface of the spiracular cartilage as well as the dor-sal aspect. However, in both Dasyatis and Uroba-tis, an additional “ventral extension” of the SP(Miyake et al., 1992) extends across the entire sur-face of the spiracular cartilage to insert on thedorsal face of the upper and lower jaws (Figs. 9g,hand 10g,h). The SP in Rhinoptera originates onthe lateral aspect of the chondrocranium, ventralto the postorbital process (Fig. 5). It then insertsvia two muscle heads, on the antero-dorsal regionof the hyomandibular cartilage and anteriorly onthe PQ. One of these two muscle heads in Rhinop-tera may be homologous to Miyake’s (1992)“ventral extension” of the SP.

AMMe. The AMMe is absent in Rhinobatosand is only present along the lower jaw in Raja(Fig. 10a) and Gymnura (Fig. 10c), originating onthe upper jaw at the corner of the mouth via con-nective tissue. In Rhinoptera (Fig. 10f), Aetobatus(Fig. 10d), Dasyatis (Fig. 10e), and Urobatis (Fig.10h) the AMMe makes a complete circuit aroundthe corner of the mouth. In Aetobatus specifically,the AMMe is parallel fibered, elongate, and wrapsaround the mouth opening, originating just ante-rior to the upper tooth row and inserting on theMK just posterior to the most procumbent (oldest)tooth row (Fig. 10d). This muscle is typically asso-ciated with the labial cartilages. In Myliobatis(Fig. 10e), this muscle does not continue its entirecircuit around the corners of the mouth and isinstead divided into two divisions (anterior andposterior) connected by a weak fibrous connection.In addition, the labial cartilages in Myliobatis are



conspicuously reduced, compared to the conditionseen in Rhinoptera and Aetobatus (Fig. 10d–f).

AMLa. The AMLa complex is typically treatedas two divisions, (named AML1 and AML2,Gonzalez-Isais, 2003; Nishida, 1990; Wilga andMotta, 1998) which we refer to as the AMLaproper (AML1) and AMMa (AML2). We observethat the AMLa shares a common tendinous inser-tion with the SB on the MK, while the AMMainsertion is restricted to the lateral corners of thelower jaw or the base of the wing process (in dur-ophagous myliobatiforms). In Raja (Fig. 10a), Rhi-nobatos (Fig. 10b) and Gymnura (Fig. 10c) theAMLa is smaller (compared to other taxa sur-veyed), originating on the anterior edge of the PQand inserting on the ventral face of the MK. InAetobatus (Figs. 9 and 10d) and Myliobatis (Figs. 9and 10e), differentiation between the AMLa,AMMa, and SB can be difficult due to the mannerin which all these muscles seem to attach in paral-lel (see AMMa description below), although thediffering fiber direction of these muscles areclearly identifiable. The AMLa and SB insert via ashared tendon on the MK, with the SB insertionslightly anterior to that of the AM lateralis. TheAMLa in Aetobatus and Myliobatis also insertsaponeurotically along the ventral face of the MK(Fig. 9d,e). The AMLa in Rhinoptera (Figs. 9 and10f) is closely associated with both the SB andAMMa, separated from the former by the trigemi-nal nerve and the latter via the AMMa tendinousconnection to the upper jaw (which the AMLaoverlays). The AMLa in Rhinoptera originates justanterior to the PQ’s jaw joint attachment site andthe posterior portion of the SB and then inserts onthe MK via a tendon it shares with the SB. InDasyatis (Figs. 9 and 10g) and Urobatis (Figs. 9and 10h), the AMLa is a pennate muscle whichoriginates on the dorsal side of the PQ (on theantero-lateral processes) and then inserts on thelateral, anterior-most process on the ventral sideof the MK. It is worth noting that the AMLa, SB,and AMMa are joined by several layers of connec-tive tissue.

AMMa. The AMMa is by far the largest jawadductor in all species surveyed. The angle atwhich each muscle division’s myofibers attach tothese fascia is reminiscent of a pennate-fiberedmuscle or several muscles working in parallel.More research is needed to investigate if thesethree divisions act as a concerted muscular unit.In Raja (Figs. 9 and 10a), Rhinobatos (Figs. 9 and10b), Gymnura (Figs. 9 and 10c), Dasyatis (Figs. 9and 10g), and Urobatis (Figs. 9 and 10h), theAMMa is a pennate-fibered muscle which coversthe entire jaw joint region. In nondurophagousstingrays, the AMMa originates on the anteriorlateral curvature of the PQ and inserts via tendonon the ventral posterior region of the MK. In Rhi-noptera, Aetobatus, and Myliobatis the AMMa is

massive, generally larger than all other jawadductors combined; it is a pennate-fibered musclewhich originates on the PQ via stout tendons onboth sides of the mouth and wraps around and upunder the “chin” of the animal’s lower jaw in alarge “U” shape (Figs. 8–10d–f). In Rhinoptera(Figs. 9 and 10f) the AMMa fits intimately withinthe “shelf” created by the wing process of the MK(Figs. 8f and 10f). In Aetobatus (Figs. 8d, 9, and10d) and to a lesser extent Myliobatis (Figs. 8e, 9,and 10e), and Rhinoptera (Figs. 9 and 10f), a por-tion of AMMa also originates aponeurotically onthe dorsal lateral surface of the PQ far anteriorlyof the tendinous insertion. The AMMa tendonswrap around the lateral labial region of the MKvia a stout fibrocartilaginous pad which has beenposited to redirect AMMa contractile force to anantero-posterior direction more in line with occlu-sion (see Summers, 2000; Summers et al., 2003).In Aetobatus and Myliobatis (Figs. 9 and 10e,f),the AMMa insertion on the posteroventral surfaceof the MK is similar to Rhinoptera, although themuscle actually covers the comparatively reducedwing process in the former two taxa (Fig. 10e,f).

AMD. The AMD, a parallel-fibered muscle, liesmedial and slightly dorsal to the tendinous regionof the AMMa, originating partly in a fossa justventral to the jaw joint and separated from theAMMa via myosepta. The AMD is presumablyabsent or far reduced in Raja, Rhinobatos, Gym-nura, Dasyatis, and Urobatis. In Rhinoptera theAMD originates and inserts entirely within thefossa created by the gap between the jaws antero-posteriorly and the jaw joint dorsally (Figs. 8f and10f). The AMD inserts on the MK within the lowerjaw’s region of the same fossa, also just ventral tothe jaw joint. In Aetobatus (Figs. 8d, 9, and 10d)and Myliobatis (Figs. 8e, 9, and 10e), the AMDoriginates far anteriorly on the PQ on a narrowskeletal projection, wrapping around the entire(dorsal, lateral, and ventral) surface of this projec-tion (Fig. 8e). This muscle may be an interior divi-sion of the AMMa complex in Rhinoptera, but isstrongly independent in Myliobatis and Aetobatus.In both Myliobatis and Aetobatus, the AMDextends far posterior and laterally from the otheradductor divisions and is considerably greater insize than in Rhinoptera. The AMD in Myliobatis isparticularly large and articulates via connectivetissue to the PCM of the propterygium (Figs. 9and 10e).

AMLi. The AMLi, is present in Gymnura andRhinoptera, is found medially, on the dorsal (lin-gual) surface of the jaws and jaw adductor bundle(facing the chondrocranium) and extends acrossboth the upper and lower jaws, adjacent (medial)to the jaw joint. The AMLi is a dumbbell-shaped,parallel-fibered muscle apparently split into twodivisions closely associated with the ligamentsholding the dorsal surfaces of the upper and lower



jaws together in articulation. This muscle isreduced or absent in Raja, Rhinobatos, Aetobatus,Myliobatis, Dasyatis and Urobatis. The AMLi orig-inates on the rostral-most, dorsal point of the PQand inserts upon the posterior-most lateral regionof the MK and is partly covered by the CHYM inRhinoptera (Fig. 10f). In Gymnura (Fig. 9c), theAMLi is smaller and wraps around the dorsal (lin-gual) surface of the lower jaws along the cornersof the mouth, medial to the jaw joint. The arrange-ment and orientation of the AMLi in Rhinopteramay suggest this muscle division is a dorsal exten-sion of the AMD, perhaps given the more dorso-ventrally compressed jaws in Rhinoptera comparedto sister myliobatids.

DISCUSSION

The relationship between muscle morphologicaldiversity, jaw suspension, and ecological niche islargely unknown in batoids. We expected that dueto biomechanical constraints on skeletal materialsand performance, durophagous stingrays willbroadly resemble (in terms of jaw muscle hypertro-phy, molariform teeth, etc.) other durophagoustaxa, like horn sharks (Huber et al., 2005; Kol-mann and Huber, 2009). Given patterns of mor-phological evolution in chondrichthyans, namelythat jaw suspension modes are fixed at deep phylo-genetic nodes and, independent lineages eachoccupy an array of trophic niches, we expectedthat muscle variation would follow patterns ofskeletal evolution. That is, mandibular muscula-ture will be remarkably disparate both within andbetween lineages, while hyoid musculature willshow disparity only between lineages (Table 2).

Durophagy in Myliobatiform Stingrays

Our data show several consistent differencesbetween durophagous stingrays and other batoidspecies apparent in the arrangement of the feed-ing musculature, particularly in the larger sizeand number of jaw adductors, a reduction in thesize of nonjaw adducting muscles, and a potentialincrease in jaw joint stabilization across duropha-gous rays. The most obvious similarity betweendurophagous rays and other durophagous elasmo-branchs (Huber et al., 2005; Kolmann and Huber,2009) is that the jaw adductor musculature inthese taxa are noticeably hypertrophied comparedto species which do not consume hard-shelled prey.Pennate-fibered muscles are more prevalent indurophagous taxa (Huber et al., 2008), whichincrease force generation over fusiform muscula-ture, while conserving volume (Cochran, 1982).

Although not restricted to durophagous rays,but particularly interesting within batoids is themanner in which some muscles insert over themaximum available surface area of the jaw skele-ton. In Dasyatis and Urobatis (as well as the elec-

tric ray, Narcine, see Dean and Motta, 2004 a,b),the primary jaw adductor muscle originates on thedorsal surface of the jaw structure, usually cover-ing the jaw joint, to wrap anteriorly over and thensharply posterior to insert on the lower jaw. InAetobatus and Myliobatis this is exemplified bythe condition of the AMD, which extends far ante-riorly to wrap around an anterior projection of thePQ (Figs. 8–10d,e). Another functional interpreta-tion of this arrangement may simply be to displacemuscle action as far from the fulcrum (jaw joint inthis case) as possible, increasing the speed atwhich the jaws can close (Fig. 2). In Myliobatisand Rhinoptera we also see an expansion of thewing process of the MK (lower jaw) associatedwith the increase in size of the AMMa (Figs. 8–10e,f). However, no other taxa exhibit the “chin-strap” morphology of the primary jaw adductor(AMMa) as seen in durophagous stingrays. In dur-ophagous sharks, jaw adductor muscle mass isusually increased by expanding the cross-sectionalarea of the quadratomandibularis (homologous toadductor mandibulae) muscles (Summers et al.,2004; Kolmann and Huber, 2009; Habegger et al.,2012). In contrast, the batoid bauplan may con-strain the available area for muscle hypertrophy;lateral expansion of the jaw adductors may beinhibited in batoids owing to the anteriorlydirected expansion of the pectoral fins and bynecessity, the encircling pectoral propterygium(Figs. 4 and 6–10). The deepening of the cranialregion in durophagous stingrays when comparedto their epibenthic relatives presumably “makesroom” for larger jaw adductors, which may alsoconstrain the size of other cranial muscles notinvolved in jaw adduction (see below).

Greater emphasis of the AMMa for jaw adduc-tion is exemplified in durophagous rays, where amuscle–tendon complex is thought to orient theAMMa muscle force in a plane perpendicular tothe dental occlusal surface (see Summers, 2000;Summers et al., 2003; Kolmann, 2012). Also,appropriation of the SB (SB) from an upper jawretracting muscle to a jaw adductor in duropha-gous taxa could potentially be associated withmaximizing the overall muscle force available forforceful biting (Figs. 2 and 8–10d–f). Expansion ofthe muscle coverage overlying the jaw joint in Rhi-noptera is also notable, mostly in the AMLa andAMMa (Figs. 8–10d–f). The AMLi and the AMDare found only in durophagous taxa. Both musclesflank (dorsally and ventrally) the jaw joint region(Fig. 2). This may aid in increasing joint stabilityby resisting tensile loading of the jaw joint, asmight occur during biting in a class 2 lever systemas proposed by Summers’ (2000) “nutcracker”model of rhinopterid jaw mechanics. Thesemuscles may also keep the jaws aligned duringfeeding as the jaw joints are conspicuously slack,exhibiting a high range of motion when the



surrounding jaw musculature is removed(Summers, 2000, Figs. 9f and 10f).

Differences between durophagous and nondur-ophagous taxa are also evident outside the man-dibular musculature. Given the changed role ofthe SB in durophagous rays from jaw suspensionto jaw adduction, it is perhaps no surprise thenthat the LP in durophagous stingrays are rela-tively immense (Figs. 8–10d–f). Together, the SPand the LP may have rendered the jaw elevatingrole of the SB (Fig. 2) redundant, allowing the SBto shift to a purely adductive function, anotherexample of interplay between jaw adductor andother cranial muscles. The DHYM in Aetobatus,Myliobatis, and Rhinoptera (Figs. 6–8d–f) are com-parably smaller than what is seen in nonduropha-gous rays (Figs. 6–8a–c,g,h), and in Rhinoptera(compared to Myliobatis and Aetobatus) theDHYM has shifted to a pennate-fiber morphology(Figs. 6–8d–f), which suggests a reduction in mus-cle volume without a decrease in performance(Cochran, 1982; Huber et al., 2008). In contrast,the size and position of the DHYM in nonduropha-gous stingrays (Figs. 6–8a–c,g,h) suggests that inaddition to depressing the hyomandibular carti-lages, the DHYM may also adduct those cartilagesand perhaps even the jaws (Fig. 2). In this posi-tion, the DHYM is capable of constricting the floorof the pharynx, depressing the hyomandibular car-tilages, and possibly aiding in jaw protrusion viamedial compression of the right and left halves ofthe jaws about their flexible symphyses (as seen innarcinid electric rays—Dean and Motta, 2004a;Dean et al., 2008). As the jaw symphyses are fusedin durophagous rays (Figs. 8–10d–f), the DHYMmay have lost some of its functional repertoire,and correspondingly, durophagous taxa are theonly stingrays in which the DHYM does not attachto the lower jaw as well as the hyomandibulae.Taken together, this suggests that the high degreeof hyoid-driven jaw protrusion seen in nonduroph-agous stingrays is impossible for durophagousrays, although rapid, cyclical jaw movements arestill a major function of feeding/excavation of prey(Sasko et al., 2006; Collins et al., 2007). Duropha-gous stingrays in particular offer a fascinatingevolutionary system in which to test biomechani-cal hypotheses regarding musculoskeletal functionand ecomorphology, and for investigating ontoge-netic trends in feeding performance.

In batoids we observed a number of musclesthat attach to the jaws via direct tendinous inser-tions as well as aponeurotic (tendinous) sheets.Aponeurotic muscle attachments (as opposed totendinous point insertions) are prevalent in chon-drichthyans, perhaps related to the relatively plia-ble nature of the skeletal cartilage relative to bonyskeletons (Summers and Koob, 2002; Summerset al., 2003). In taxa with aponeurotic insertions,the muscle attachment is integrated broadly into

the fibrous outer perichondrium that wraps theskeleton, thereby distributing strain on the skele-ton over a larger area. In durophagous taxa, how-ever, a more robust jaw skeleton (stiffened bycortical thickening and trabecular internal rein-forcement; Summers, 2000) coupled with fusion ofcertain skeletal elements (i.e., jaw symphyses)makes aponeurotic insertions unnecessary. Thesedirect tendinous insertions are more amenable toefficient force transfer (Dean et al., 2007b) acrossa rigid (jaw) structure and reduce the area neededfor muscles to attach. Analysis of correlationsamong muscle activity, skeletal strains during nat-ural feeding behaviors, and muscle attachmentmorphologies in these animals will help to eluci-date basic form-function relationships in elasmo-branch skeletal anatomy, which remain woefullyunderstudied.

Conservation and Variation inChondrichthyan Feeding Musculature

In general, differences in the cranial anatomy ofchondrichthyan fishes are primarily observed inthe muscles stemming from the hyoid and mandib-ular embryonic muscle plates (Miyake et al., 1992,Table 2). The evolutionary malleability of thesetwo embryonic plates has played a large role indriving ecological diversity at different points inthe diversification of cartilaginous fishes. Broadcomparisons across the major groups of chon-drichthyans —sharks, batoids and holocepha-lans— illustrates that muscle variability withinthese taxa is explained by both developmentalidentity and functionality (Miyake et al., 1992).Phylogenetic differences in the arrangement ofmuscles stemming from the branchial muscle plateare not as apparent as in the mandibular andhyoid plates (Table 2); therefore we restrict ourbrief discussion to the latter two muscle groupsand the hypaxials (see Miyake et al., 1992 foradditional information).

The hyoid muscle plate in chondrichthyan fishesis overwhelmingly disparate between batoids andsharks, with even more overt differences betweenbatoids and holocephalans (Table 2, Miyake et al.,1992; Anderson, 2008). This is presumably associ-ated with the increasing decoupling between thejaws and chondrocranium (mediated by the hyo-mandibular cartilages, Fig. 1), and specific tobatoids, the ventral orientation of the mouth. Hol-ocephalans are characterized by fusion of theupper jaw to the cranium (holostyly—Fig. 1a).Sharks exhibit various configurations of pre andpostorbital, as well as hyostylic (Fig. 1b,d) connec-tions between the jaws, hyomandibulae, and thecranium. Finally, batoids exclusively are charac-terized by “euhyostylic” suspension (or an“unsuspended” PQ—Maisey, 1980; Fig. 1c). Most ofthe differences in the overall arrangement of the



hyoid muscle plate seem to be restricted to crownclades. Between squaloid and galeoid sharks, forexample, hyoid muscle arrangement (Wilga, 2005;Soares and de Carvalho, 2013) is conserved,despite these two clades having evolved independ-ently from each other (�200 mya) for nearly aslong as sharks have been separated from batoids(Aschliman et al., 2012). Two of the muscles aris-ing from the hyoid plate in batoids (LHYM andDHYM, Figs. 5 and 6) are involved with jaw sus-pension and another two (DR and LR, Figs. 5 and6) deal with manipulation of the rostral region insome batoids (skates and guitarfish, Table 1).Sharks do not have flexible snouts, but do exhibithighly kinetic jaws, whereas holocephalans haveneither flexible snouts nor kinetic jaws (Maisey,1980), making the absence of three of the fourabove muscles in nonbatoids unsurprising. Thismay suggest that further decoupling of the hyoidhas allowed muscle divisions to acquire novel func-tionality beyond hyoid movement, as is the casefor the depressor and LR, which position the snoutin batoids.

Our data support the hypothesis that jaw sus-pension musculature, classified here as muscleswhich adduct the hyomandibular cartilages towardthe cranium, are conserved across elasmobranchs(Table 2). During ventilation and feeding theDHYM and CHYM are oriented to depress anddraw the hyomandibulae medially to expand thepharynx (Fig. 2). LHYM works as an antagonist tothe DHYM and CHYM, drawing the hyomandibu-lar cartilages dorsally, thus returning the hyoman-dibulae to a resting position (Fig. 2). In sharks,which lack the CHYM (Miyake et al., 1992; Ander-son et al., 2008), depression of the hyomandibularcartilages are modulated solely by the CH. Move-ment of the hyoid then is always actuated by botha hyoid and hypaxial muscle component, withmore divisions used in batoids (two hypaxial andtwo hyoid muscles, Fig. 2) than sharks (one hypax-ial and one hyoid muscle, Table 2). The rectus cer-vicus (first embryonic hypaxial precursor) givesrise to the CH, and CARC muscles in sharks, inaddition to the CHYM in batoids (Table 2). Thegenio-hyoideus (another embryonic hypaxial pre-cursor) gives rise to the CM in all taxa. The rectuscervicus is undifferentiated in ratfishes (Didier,1995, Table 2). Given the increasingly hyostylicnature of the jaws in batoids when compared toratfish, and to a lesser extent, sharks, the CHYMmay have evolved to maintain control and stabilityin this comparatively “free” jaw suspension.

In terms of “total” jaw suspension, involvingboth suspension of the hyoid AND the jaws, man-dibular plate muscles are also implicated (Tables 1and 2). The SP, with its association in some taxawith the hyomandibular cartilages (durophagousrays) and the upper jaw (Dasyatis and Urobatis),assist in jaw suspension in addition to regulating

the spiracular aperture. In all batoids, the LP isinvolved with retracting the jaw after the expan-sive phase of feeding (Wilga and Motta, 1998). TheLP and SP (and the SB in some taxa) are alwayspresent as separate divisions in batoids, aiding inelevation of the PQ (Miyake et al., 1992). Thesemuscles are variably expressed in sharks, depend-ing on the presence or absence of the spiracle inthese clades (Huber et al., 2011; Soares and deCarvalho, 2013); however, only the LP is found inratfishes (Didier, 1995; Huber et al., 2011, Table2).

Further variation in mandibular muscles isobvious between lineages and within clades. Thelevator series of muscles in ratfish (levator mandi-bulae, levator cartilaginis, and levator anguli andlabialis muscles) stem from the mandibular muscleplate and presumably aid in movement of thelabial cartilages (Didier, 1995; Huber et al., 2011).The specific relationship of these muscles to thosein batoids and sharks is poorly known, besidesgeneral developmental homology (mandibularplate, Miyake et al., 1992, Table 2). In holocepha-lans, the jaw adductors consist of only two divi-sions, the AM anterior and posterior (Huber et al.,2008, 2011, Table 2). The quadratomandibularis insharks is generally divided into four or five majordivisions, each with slight differences in musclefiber direction and architecture (Motta and Wilga,1995, 1999; Wilga and Motta, 1998; Huber andMotta, 2004; Soares and de Carvalho, 2013, Table2). The adductor mandibulae complex in batoids isdivided in a similar manner into multiple divisions(AMMe, AMLa, AMMa, AMD, and AMLi). How-ever, these muscles in batoids may serve a greatervariety of functions related to feeding. The AMMepresumably functions in conjunction with thelabial cartilages during suction feeding to manipu-late the corners of the mouth (Dean and Motta,2004 a,b). We suggest the AMD and AMLi in Rhi-noptera and Myliobatis may help to stabilize thejaw joint during forceful biting, especially whenunilateral adductor activity places the contra-lateral jaw joint in tension (Summers, 2000).Lastly, the AMLa and AMMa serve as the primaryjaw adductors (Kolmann, 2012). In both elasmo-branch clades, the SB (or preorbitalis, Mallat,1997) is involved in lower jaw adduction andupper jaw protrusion (Wilga and Motta, 1998;Huber et al., 2011).

CONCLUSIONS

Our evolutionary understanding of jawed verte-brates is predicated within the context of neonto-logical cranial anatomy. In particular,understanding current anatomical, functional andecological diversity in organisms like chon-drichthyans gives us perspective on ancient radia-tions and their legacy. The considerable degree of



variation in muscle position, muscle size and prev-alence of novel muscle divisions is remarkablegiven the relatively low species diversity of batoidfishes and other elasmobranchs when compared toteleosts. We believe that the euhyostylic jaw sus-pension plays a considerable role in explainingtrophic diversity in batoid fishes, analogous to howskeletal decoupling has been proposed to havefacilitated functional, morphological, and ecologicaldiversification in some teleosts (Schaefer andLauder, 1996; Hulsey et al., 2006; Lujan andArmbruster, 2012). The comparatively “free” jawsuspension of batoids seems to have providedopportunities for some muscles (e.g., hyoid andmandibular muscles) to contribute to new behav-iors while other muscles (e.g., branchial muscles)remain relatively unchanged. Our data suggestthen, among derived stingrays at least, decreasingskeletal support for the feeding apparatus and anincrease in muscular control and stability. Thistendency in batoids is illustrated by muscles origi-nating and then inserting on other muscles, theprevalence of novel or subdivided muscles, as wellas the frequency with which direct tendon attach-ment to the skeleton are supplemented by apo-neurotic sheets. The numerous adductorsubdivisions in the jaws of durophagous stingrays,some of which overlie a conspicuously slack jawjoint ligament, are particularly suggestive of reli-ance on muscular, rather than skeletal supportand control.

Repurposing of certain muscles for novel roles aswe have described in batoids is probably not atypi-cal for vertebrates, however, we document thesenovelties and their potential functional ramifica-tions in a clade that has not received much atten-tion from evolutionary or ecological studies. Weconclude by stating several functional-evolutionaryhypotheses, which need further testing:

1. Less direct skeletal articulation in jaw suspen-sion promotes greater functional diversity,either caused by or related to a release of con-straints from developmental associations. Fol-lowing this, we would expect the mandibularand hyoid muscles to exhibit greater functionaland behavioral variability than hypaxial and inparticular, branchial musculature.

2. Less direct skeletal articulation in jaw suspen-sion facilitates greater muscular morphologicalcomplexity (in terms of muscle division multi-plicity, novelty, method of attachment,pennation).

3. The more compliant skeleton in cartilaginousfish necessitates direct muscle insertions sup-plemented by aponeuroses, which perhaps miti-gates point-loading strains.

These hypotheses require experimental valida-tion at the level of genotype expression, biome-chanical and physiological traits, as well as

behavioral kinematics. Updated studies of embryo-logical development through tissue density-specificstaining methods is recommended to confirm mus-cle developmental homology across chon-drichthyans and other fishes considering bothphylogeny and ontogeny. To explore the functionaldiversity of cranial muscles, kinematic evaluationof feeding and respiration behaviors need to beexamined. Electromyography methods document-ing muscle activity during kinematic events inchondrichthyan fishes (as has been performed to amuch broader degree in sharks; see Wilga et al.,2012) are needed to confirm whether musclebehavioral variability follows morphological andphysiological variability. Taken together, theseavenues of research would build off of previousand current anatomical study, providing an inte-grated direction for investigating form-functionrelationships in this understudied, yet trophicallydiverse group of fishes.

AUTHOR CONTRIBUTIONS

M.A.K., R.D.G., and D.R.H procured specimensand funding. M.A.K., M.N.D. and D.R.H. per-formed dissections and exploratory muscle descrip-tions. M.A.K., D.R.H., R.D.G. and M.N.D. wrotethe bulk of the article. The authors declare no con-flict of interest in the publication of thissubmission.

ACKNOWLEDGMENTS

The authors thank C. Bedore (Florida AtlanticUniversity), Dana Bethea (National Marine Fish-eries Service), S. Mulvany (University of SouthFlorida), K. Parsons, C.J. Sweetman, and R.Fisher (Virginia Institute of Marine Science), J.Christofferson and G. Poulakis (Florida Fish andWildlife Conservation Commission) for assistancewith specimen collection. G. Erickson, S. Steppan(Florida State University), and two anonymousreviewers provided much-appreciated advice oninitial drafts. Special thanks go to the FloridaState Coastal and Marine Laboratory (FSUCML)for providing an excellent facility for research andlearning.

LITERATURE CITED

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

Aschliman NC, Nishida M, Inoue JG, Rosana KM, Naylor GJP.2012. Body plan convergence in the evolution of skates and rays(Chondrichthyes: Batoidea). Mol Phylogenet Evol 63:28–42.

Bigelow HB, Schroeder WC. 1953. Sharks, sawfishes, guitarfishes,skates and rays. Chimaeroids. In: Tee-Van J, Breder CM, Hilde-brand SF, Parr AE, and Schroeder, We (eds.). Fishes of the West-ern North Atlantic. Part 2. Sears Foundation for MarineResearch, New Haven: Yale Univ. pp. 1–514.

Claeson KM. 2011. The synarcual cartilage of batoids withemphasis on the synarcual of Rajidae. J Morphol 12:1444–1463.



Cochran GVB. 1982. A Primer of Orthopaedic Biomechanics.New York, NY: Churchill Livingstone.

Collins AB, Heupel MR, Hueter RE, Motta PJ. 2007. Hard preyspecialists or opportunistic generalists? An examination ofthe diet of the cownose ray, Rhinoptera bonasus. Mar Fresh-water Res 58:135–144.

Dean MN, Motta PJ. 2004a. Anatomy and functional morphologyof the feeding apparatus of the lesser electric ray, Narcine bra-siliensis (Elasmobranchii: Batoidea). J Morphol 262:462–483.

Dean MN, Motta PJ. 2004b. Feeding behavior and kinematicsof the lesser electric ray, Narcine brasiliensis. Zoology 107:171–189.

Dean MN, Wilga CD, Summers AP. 2005. Eating without hands ortongue: Specialization, elaboration and the evolution of prey proc-essing mechanisms in cartilaginous fishes. Biol Lett 1:357–361.

Dean MN, Bizzarro JJ, Summers AP. 2007a. The evolution ofcranial design, diet, and feeding mechanisms in batoid fishes.Int Comp Biol 47:70–81.

Dean MN, Azizi E, Summers, AP. 2007b. Uniform strain inbroad muscles: Active and passive effects of the twisted ten-don of the spotted ratfish Hydrolagus colliei. J Exp Biol 210:3395–3406.

Dean MN, Ramsay JB, Schaefer JT. 2008. Tooth reorientationaffects tooth function during prey processing and tooth ontog-eny in the lesser electric ray, Narcine brasiliensis. Zoology111:123–134.

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

Douady CJ, Dosay M, Shivji MS, Stanhope MJ. 2003. Molecularphylogenetic evidence refuting the hypothesis of Batoidea(rays and skates) as derived sharks. Mol Phylogenet Evol 26:215–221.

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

Gonzalez-Isais M. 2003. Anatomical comparison of the cephalicmusculature of some members of the Superfamily Myliobatoi-dea (Chondrichthyes): Implications for evolutionary under-standing. Anat Rec 271:259–272.

Gonzalez-Isais M, Dominguez HMM. 2004. Comparative anat-omy of the Superfamily Myliobatoidea (Chondrichthyes) withsome comments on phylogeny. J Morphol 262:517–535.

Habegger ML, Motta PJ, Huber DR, Dean MN. 2012. Feeding bio-mechanics and theoretical calculations of bite force in bull sharks(Carcharhinus leucas) during ontogeny. Zoology 115:354–364.

Huber DR, Motta PJ. 2004. Comparative analysis of methodsfor determining bite force in the spiny dogfish Squalus acan-thias. J Exp Zool 301:26–37.

Huber DR, Eason TG, Hueter RE, Motta PJ. 2005. Analysis ofthe bite force and mechanical design of the feeding mecha-nism of the durophagous horn shark Heterodontus francisci.J Exp Biol 208:3553–3571.

Huber DR, Dean MN, Summers AP. 2008. Hard prey, soft jawsand the ontogeny of feeding mechanics in the spotted ratfish,Hydrolagus colliei. J R Soc Int 5:941–952.

Huber DR, Soares MC, de Carvalho MR. 2011. CartilaginousFishes Cranial Muscles. In: Farrell AP, editor. Encyclopediaof Fish Physiology: From Genome to Environment, Vol. 1.San Diego, CA: Academic Press. pp 449–462.

Hulsey CD, Garcia de Leon FJ, Rodiles-Hernandez R. 2006.Micro- and macroevolutionary decoupling of cichlid jaws: A testof Liem’s key innovation hypothesis. Evolution 60:2096–2109.

Kolmann MA. 2012. Ecomorphological consequences of the feed-ing mechanism in the cownose ray, Rhinoptera bonasus [The-sis]. Tallahassee, FL: Florida State University. pp 76.

Kolmann MA, Huber DR. 2009. Scaling of feeding biomechanicsin the horn shark Heterodontus francisci: Ontogenetic con-straints on durophagy. Zoology 112:351–361.

Lovejoy NR. 1996. Systematics of myliobatid elasmobranchs:With emphasis on the phylogeny and historical biogeographyof neotropical freshwater stingrays (Potamotrygonidae: Raji-formes). Zool J Linn Soc 117:207–257.

Lujan NK, Armbruster JW. 2012. Morphological and functionaldiversity of the mandible in suckermouth armored catfishes(Siluriformes: Loricariidae). J Morphol 278:24–39.

Maisey JG. 1980. An evaluation of jaw suspension in sharks.Am Mus Novit 2706:1–17.

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

Mallat J. 1997. Shark pharyngeal muscles and early vertebrateevolution. Acta Zool 78:279–294.

Mara KR, Motta PJ, Huber DR. 2009. Bite force and perform-ance in the durophagous bonnethead shark, Sphyrna tiburo.J Exp Zool 311:1–11.

Miyake T, McEachern JD, Hall BK. 1992. Edgeworth’s legacy ofcranial muscle development with an analysis of muscles inthe ventral gill arch region of batoid fishes (Chondrichthyes:Batoidea). J Morphol 212:213–256.

Motta PJ, Huber DR. 2012. Prey capture behavior and feedingmechanics of elasmobranchs. In: Carrier JC, Musick JA, Hei-thaus MR, editors. The Biology of Sharks and their Relatives,2nd ed. Boca Raton, FL: CRC Press LLC. pp 153–209.

Motta PJ, Wilga CAD. 1995. Anatomy of the feeding apparatusof the lemon shark, Negaprion brevirostris. J Morphol 226:309–329.

Motta PJ, Wilga CAD. 1999. Anatomy of the feeding apparatus ofthe nurse shark, Ginglymostoma cirratum. J Morphol 241:33–60.

Mulvany S, Motta PJ. 2013. The morphology of the cephaliclobes and anterior pectoral fins in six species of batoids. JMorphol 274:1070–1083.

Nishida K. 1990. Phylogeny of the Suborder Myliobatoidei.Memoirs of the Faculty of Fisheries, Hokkaido University.Japan: Hokkaido University Press.

Notarbartolo-di-Sciara G. 1987. Natural history of the rays of theGenus Mobula in the Gulf of California. Fishery Bull 86:45–66.

Sasko DE, Dean MN, Motta PJ, Hueter RE. 2006. Prey capturebehaviour and kinematics of the Atlantic cownose ray, Rhi-noptera bonasus. Zoology 109:171–181.

Schaefer SA, Lauder GV. 1996. Testing historical hypotheses ofmorphological change: Biomechanical decoupling in loricar-ioid catfishes. Evolution 50:1661–1675.

Soares MC, de Carvalho MR. 2013. Mandibular and hyoidmuscles of galeomorph sharks (Chondrichthyes: Elasmobran-chii), with remarks on their phylogenetic intrarelationships. JMorphol 274:1111–1123.

Summers AP. 2000. Stiffening the stingray skeleton—An inves-tigation of durophagy in myliobatid stingrays (Chondrich-thyes, Batoidea, Myliobatidae). J Morphol 243:113–126.

Summers AP, Koob TJ. 2002. The evolution of tendon-morphology and material properties. Comp Biochem PhysiolA 133:1159–1170.

Summers AP, Koob-Emunds MM, Kaijura SM, Koob TJ. 2003.A novel fibrocartilaginous tendon from an elasmobranch fish(Rhinoptera bonasus). Cell Tissue Res 312:221–227.

Summers AP, Ketcham RA, Rowe T. 2004. Structure and func-tion of the horn shark (Heterodontus francisci) craniumthough ontogeny: Development of a hard prey specialist. JMorphol 260:1–12.

Wilga CD. 2002. A functional analysis of jaw suspension inelasmobranchs. Biol J Linn Soc 75:483–502.

Wilga CD. 2005. Morphology and evolution of the jaw suspen-sion in lamniform sharks. J Morphol 265:102–119.

Wilga CD, Motta PJ. 1998. Feeding mechanism of the Atlanticguitarfish Rhinobatos lentiginosus: Modulation of kinematicsand motor activity. J Exp Biol 201:3167–3184.

Wilga CD, Motta PJ. 2000. Durophagy in sharks: Feedingmechanics of the hammerhead Sphyrna tiburo. J Exp Biol203:2781–2796.

Wilga CD, Maia A, Nauwelaerts S, Lauder GV. 2012. Prey han-dling using whole-body fluid dynamics in batoids. Zoology115:47–57.



Documents

Myological variability in a decoupled skeletal system: Batoid cranial anatomyutweb.ut.edu/hosted/faculty/dhuber/images/Kolmann et al... · 2016-03-03 · Myological Variability in