Pharyngeal biting mechanics in centrarchid and cichlid ...frietsongalis.nl/wp-content/uploads/2009/06/Galis-Drucker-1996-J...9516, NL-2300 RA Lriden, ... Pharyngeal biting in perciform

J. Evol. Bid. 9: 641 670 (1996) IOIO-06lX/96/OSO641 30$ l.50+0.20/0 c 1996 Birkhauser Verlag. Bad

Pharyngeal biting mechanics in centrarchid and cichlid fishes: insights into a key evolutionary innovation

F. Galis’.* and E. G. Drucker’

‘Institute ,for Evolutioncrry and Ecological Sciences, Univrrsity of’ Leidtw, P.O. Box 9516, NL-2300 RA Lriden, The Nrtlzrrlunds, c-mud: gulis ((I rul.~~fb.leic~enunic.nl ‘Museum of Comparative Zoology, Horvd Uniwrsit)~, Cumhridge, MA 02138, USA

Key n~r&s: Cichlidae; Centrarchidae; mechanical model; adaptive radiation; diversity; evolution.

Abstract

This study compares the pharyngeal biting mechanism of the Cichlidae, a family of perciform fishes that is characterized by many anatomical specializations, with that of the Centrarchidae, a family that possesses the generalized perciform anatomy. Our objective was to trace the key structural and functional changes in the pharyngeal jaw apparatus that have arisen in the evolution from the generalized to derived (cichlid) perciform condition.

We propose a mechanical model of pharyngeal biting in the Centrarchidae and compare this with an already existing model for pharyngeal biting in the family Cichlidae. Central to our centrarchid model is a structural coupling between the upper and lower pharyngeal jaws. This coupling severely limits independent movement of the pharyngeal jaws, in contrast to the situation in the speciose Cichlidae, in which the upper and lower pharyngeal jaw movements are to a large extent independent. We tested both models by electrically stimulating nine muscles of the branchial and hyoid apparatuses in three centrarchid and three cichlid species. The results confirmed the coupled movement of the upper and lower pharyngeal jaws in the Centrarchidae and the independence of these movements in the Cichlidae.

We suggest that the key structural innovation in the development of the functionally versatile cichlid (labroid) pharyngeal jaw apparatus was the decoupling of epibranchials 4 from the upper pharyngeal jaws. This structural decoupling implies

* Author for correspondence.

641

642 Galis and Drucker

the decoupling of the movements of the upper and lower pharyngeal jaws and leads to a cichlid (labroid) type of pharyngeal bite. The initial decoupling facilitated a cascade of changes, each leading to improved biting effectiveness and/or to increased mobility and mechanical flexibility of the pharyngeal jaws. The shift of insertion of the m. levator externus 4 which has been considered the primary innovation in the transformation probably arose secondarily.

The transformation of the pharyngeal biting mechanism in the perciforms is an excellent example of decoupling of structures associated with diversification of form and function and with increased speciation rates.

Introduction

The extensive diversification of labroid fishes (Teleostei: Perciformes) has been attributed to the development of a morphologically specialized and functionally versatile pharyngeal jaw apparatus (Fryer and Iles, 1972; Greenwood, 1973; Liem, 1973, 1978, 1979; Yamoaka, 1978; Liem and Sanderson, 1986; Jensen, 1990). Much attention has been paid to the structure and function of the derived pharyngeal jaw apparatus of labroids (e.g. Greenwood, 1965; Liem, 1973, 1978, 1979; Yamaoka, 1978; Hoogerhoud and Barel, 1978; Liem and Greenwood, 1981; Kaufman and Liem, 1982; Liem and Sandcrson, 1986; Stiassny and Jensen, 1987; Wainwright, 1987, 1988; Claes and DeVree, 1991 a, b; Drucker and Jensen, 1991; Vandewalle et al., 1992; Galis, 1992, 1993a, b). In contrast, comparatively little attention has been given to the mechanism of pharyngeal biting in less derived (i.e., non-labroid) perciforms retaining ancestral conditions (but see Liem, 1970, on Nandidae; Lauder, 1983a, b on Centrarchidae; Wainwright, 1989, on Haemulidae; Vandewalle et al., 1992, on Serranidae). To understand the evolutionary transformation of the perciform pharyngeal jaw apparatus and to determine possible key innovations, knowledge of both the ancestral and derived conditions is necessary.

The first objective of this study was to provide and test a mechanical model of pharyngeal biting in a perciform family that can serve as a model for the ancestral condition of the pharyngeal jaw apparatus of labroids. Percoids are a group of basal perciforms (Greenwood et al., 1966; Lauder and Liem, 1983; Stiassny and Jensen, 1987) that possesses the generalized pharyngeal muscular anatomy and that is though to be most closely related to labroids (Kaufman and Liem, 1982; Stiassny and Jensen, 1987). Therefore, we developed a model for pharyngeal biting in the generalized percoid family Centrarchidae. In part this model is based upon a model for haemulids (Wainwright, 1989). A second objective was to test a mechanical model of pharyngeal biting in the Cichlidae (Galis, 1992) a family characterized by anatomical specialisations of the pharyngeal jaw apparatus. A third objective was the comparison and interpretation of the test results of both models. In comparing the generalized perciform pharyngeal biting mechanism of the Centrarchidae with the specialized pharyngeal biting mechanism of the Cichlidae, we propose a transformation scheme of the major changes in the pharyngeal jaw apparatus in evolution from the non-labroid to labroid conditions. In this tranformation scheme

Pharyngeal biting in perciform fishcs 643

a key structural innovation, a decoupling of structures, is indicated which releases important functional constraints and allows a cascade of other changes. We examine the role of functional decoupling in evolutionary change.

Materials and methods

The pharyngeal jaw apparatus of fresh and ethanol-preserved adult specimens of three species of centrarchid (Lepomis gibbosus, L. macrochirus and Micropterus salmoides) and three species of cichlid (Labrochomis ishmaeli, Oreochromis niloticus and Cichlusoma citrinellum) were dissected for anatomical study. Osteological observations were made from skeletal preparations of L. gibbosus, M. salmoides and Lepomis gulosus (Museum of Comparative Zoology, Harvard University) and from cleared and stained specimens of Enneacanthus gloriosus, Centrarchus micropterus (centrarchids; Museum of Comparative Zoology, Harvard University) and Asta- toreorhromis alluuudi and Ptyochromis suuvugei (cichlids; University of Leiden, The Netherlands).

Centrarchids for experimentation were collected by seine in the Charles River, Cambridge, MA. Lubrochromis ishmaeli specimens were caught in Mwanza bay in Lake Victoria by the Haplochromis Ecology Survey Team of the University of Leiden. Oreochromis niloticus and Cirhlusoma citrinellum were obtained from commercial suppliers in Massachusetts. The fishes were maintained on a diet of pellets, squid, and smelt in SO- 160 1 aquaria at 22224 ‘C for the centrarchids and 24-26 “C for the cichlids.

A mechanical model of pharyngeal biting in centrarchids was developed from anatomical observations and manipulation of fresh specimens. Forces were consid- cred under the assumption of static equilibrium. The lengths of the bones and muscles were chosen according to their approximate dimensions in the animals. The model was tested by electrically stimulating individual muscles of the branchial and hyoid skeletons in live anaesthetized fishes and observing the resulting movements of the pharyngeal jaws (Tab. 1).

Stimulation experiments were carried out on five centrarchid specimens: two L. gibbosus ( 13.8 and 14.5 cm standard length), two L. macrochirus ( 12.5, 16.4 cm SL) and one M. salmoides (17.5 cm SL). In order to test the function of homologous muscles in the Cichlidae, stimulation experiments were performed on five cichlid specimens: two 0. niloticus (14.4, 14.5 cm SL), two L. ishmaeli (X.9, 9.8 cm SL) and one specimen of C. citrinellum (12.4 cm SL). Fishes were anaesthetized with a solution of tricaine methane sulfonate. The pharyngeal muscles were exposed by removing the overlying mucous epithelium. The operculum was manually ab- ducted, and in some cases removed, to ensure an unobstructed lateral view of the pharyngeal jaw apparatus. Bipolar fine-wire electrodes (0.051 mm diam.) were inserted via 27 gauge hypodermic needles into nine muscles (see Tab. 1). Approxi- mately 0.3 mm of insulation was removed from the tip of each electrode wire. The position of each pair of electrodes was verified by inspection through the translu- cent muscle tissue, or in the case of the retractor dorsalis, by post-mortem


Table 1. Action of muscles upon electrical Stimulation.

Ml&e Origin Insertion Action

UPJ LPJ

In wntrtrrchidc

LE4 LP RD PC1 PC-E PH GH SH I-R. VENTR.

In cichlirl:,

LE4 LP RD

PC1 P(‘E

PFI GH SH TR.VENTR.

Neurocranium Neurocranium Neurocranium Pectoral girdle Pectoral girdle Urohyal Dentary Pectoral girdle Ceratobr. 4

Neurocranium Neurocranium Neurocranium Pectoral girdle Pectoral girdle

Urohyal Dentary Pectoral girdle CB4

Epibranchial 4 depression. retract.* elevation, protraction

Epibranchial 4 depression, retract.* elevation, protraction

UPJ retraction retraction

LPJ retraction retraction*, elevation**

LPJ depression. retraction*

LPJ protraction* Hyoid elevation, protraction

Urohyal depression, retraction

LPJ I1.S.

LPJ LPJ/EP4 UPJ

LPJ LPJ

LPJ Hyoid Urohyal LPJ

retraction

elevation, protraction elevation, protraction

retraction depression. retraction

protraction clcvalion, protraction depression, retraction

elevation

Boldface indicates previously unreported action. indicates no movement observed: n.s. indicates muscle was not stimulated.

* Confirmation of action observed in haemulids by Wainwright (1989). ** Elevation only in L. &ho.sus. Origins and insertions for centrdrchids from Wintcrbottom ( 1974). Lauder (1983); for cichlids fi-om

Liem (1973). Ankcr ( 1978). Abbrevations: LE4: m. levator externus 4: LP: m. levator posterior; RD: m. retractor dorsalis; PCI: m. pharyngocleithralis internus; PCE: m. pharyngocleithralis externus; PH: m. pharyngohyoideus; GH, m. geniohyoideus; SH: m. sternohyoideus: CB4: ceretobranchial 4: LPJ: lower pharyngeal jaw; EB4: epibranchial 4; UPJ: upper pharyngeal ,jaw.

dissection. A muscle stimulator was used to deliver at first twitch stimuli (10 msec pulse duration, 4- 15 V) and then tetanic stimuli (40 Hz, IO msec pulse duration, 4-15 V). Movements of the pharyngeal jaws during muscle stimulation were observed through a stereomicroscope and were confirmed by observation of at least two people.

Anatomical couplings

Dissection of preserved specimens and manipulation of freshly killed specimens revealed that there are three important structural couplings among the bony elements of the pharyngeal jaw apparatus of centrarchids (Fig. 1). Two of these three couplings are modified in the Cichlidae.

Pharyngeal biting in perciform fishes 645

Fig. 1. Caudal view of the bones of the pharyngeal jaw apparatus of (A) the piscivorous centrarchid

Microprerus scdmoidrs and (B) the molluscivorous centrarchid Lc~~~ni.s gihbosus. The structural couplings are indicated with numbers I to 3 (see text). The muscles that connect epibranchials 4 with the lower pharyngeal jaws and form part of coupling 3 are not shown.

Coupling 1: Fourth epibvanrhials ~ upper phuryngeul ,@~s

In the Centrarchidae there is an articulation between the lateral face of the upper pharyngeal jaw (toothed pharyngobranchials) and the medial surface of epibranchial 4 (Fig. I: coupling l), as in haemulids (Wainwright, 1989). Following Wainwright ( 1989) we expect that upon contraction of the m. levator externus 4 or m. levator posterior, which originate on the neurocranium and insert on epibranchial 4 (Fig. 2) the epibranchials 4 pivot (see for explanation of the fixed rotation center page 655): their medial limbs are depressed and their lateral limbs are elevated (Fig. 3). The downward rotating medial limbs, in turn, depress the upper pharyngeal jaws (Figs. 1 and 3). Thus, we predict that this coupling transmits the force exerted by m. levator posterior and externus 4 on epibranchials 4 to the upper pharyngeal jaws, as in the Haemulidae (Wainwright, 1989).

In the Cichlidae coupling 1 is lost: epibranchials 4 lie against the dorsolateral surface of the upper pharyngeal jaws (Fig. 4; Bare1 et al., 1976; Stiassny, 1981), but the two elements are capable of independent movement (Galis and Terlouw, unpublished data).

The m. levator externus 4 of cichlids runs from the neurocranium to the lower pharyngeal jaw (Figs. 5 and 9; see also Liem and Greenwood, 198 1; Kaufman and Liem, 1982; Stiassny and Jensen, 1987; VandeWalle et al., 1992; VandeWalle et al., 1994) rather than to epibranchial 4 as in centrarchids and suspends the fused lower pharngeal jaws in a muscle sling (Liem, 1973, 1978). The force of this muscle is, therefore, expected to act directly upon the lower pharyngeal jaw (Figs. 5 and 6) instead of indirectly upon the upper pharyngeal jaws via rotation of epibranchials 4 (Fig. 3).

In some species the insertion of m. levator posterior has also shifted to the lower pharyngeal jaw and contributes to the muscle sling (Fig. 5; Liem, 1978) but in


Fig. 2. Lateral view of the pharyngcal jlw musculature in (A) the piscivorous centrarchid Micro~~/erus

.vrr/r~roitk~~~ and in (B) the molluscivorous centrarchid Lqx~i.s ~ihhosus.

many others it runs in the primitive condition to epibranchials 4 (Liem, 1973; Anker, 1978; Galis, 1992). In the cichlids examined in this study body configura- tions occur: in C. citvinellum and L. ishmacli the m. levator posterior continues to the lower pharyngeal jaw and in 0. niloticus it inserts on epibranchial 4. Upon contraction of the m. levator posterior in cichlids with the primitive insertion on epibranchials 4 we expect elevation of the fourth epibranchials and sliding of these elements along the upper pharyngeal jaws without pivoting (cf. Fig. 13). We expect no downward directed movement of epibranchials 4 and upper pharyngeal jaws as in centrarchids and haemulids.

Pharyngeal biting in pcrciform fishes 647

Fig. 3. Proposed mechanical model of pharyngeal biting in the Centrarchidae. (A) Schematic caudal view and (B) schematic lateral view of the pharyngeal jaw apparatus. Solid line arrows indicate the direction of the translation and rotation of the bones. Dashed line arrows indicate the direction of the forces exerted by muscles. Activity of left and right tn. levator posterior and m. lcvator cxternus 4 leads

to both depression of the upper pharyngeal jaws and to elevation of the lower pharyngeal jaws -*occlusion of the pharyngeal jaws. Upper pharyngcal jaw dcprcasion occurs via rotation of the fourth epibranchial around their centers. Dcprcssion of the medial limb of epibranchial 4 leads to dcprcssion of the upper pharyngcal jaw (Wainwright, 1989). Elevation of the lateral limb of epi-

branchials 4 results in clcvatlon of the lower pharyngeal jaw (and ceratobranchlal 4). Activity of the m. geniohyoideus also results in elevation of the lower pharyngcal ,jaws.

Fig. 4. Caudal view of the bones of the pharyngeal jaw apparatus of the cichlid Lr/hroc~/~ro/ni.r itl~n~rli. Coupling 3 bctwccn lower pharyngeal jaw consists at lcast partly of the myoseptum of m. levator externus 4 (see tat). Epibranchials 4 can slide along the upper pharyngeal J~VVS. The lower pharyngeal ,jaws in cichlids are fused into a single bone.

648 Galis and Druckcr

Fig. 5. Schematic illustration of head muscles effecting movements of the upper and lower pharyngeal ,jaw in the cichlid Luhrochromis ishmcdi. For abbreviations, see Table I. Modified from Galis ( 1993b).

Coupling 2: Fourth cerutohrunchiuls - lower pharyngeal juws

In the Centrarchidae the lower pharyngeal jaws (toothed fifth ceratobranchials) are firmly connected to the fourth ceratobranchials (Figs. 1, 7: coupling 2). There are ligamentous connections which greatly restrict independent movement of the fourth ceratobranchials and the lower pharyngeal jaws. The m. transversus ventralis anterior (Fig. 7) further limits independent movement of the fourth and fifth ceratobranchials. In centrarchids this muscle runs in the primitive perciform condition without interruption from left to right fourth ceratobranchial, forming a muscular sling beneath the lower pharyngeal jaws (Fig. 8a; Winterbottom, 1974; Wainwright, 1989; Stiassny, 1990). Contraction of this muscle thus strengthens the connection between the lower pharyngeal jaws and the ceratobranchials 4. At rest the muscle presumably also functions as a ligament. In L. gihbosus, a centrarchid which crushes hard prey (Keast, 1978; Lauder, 1983a), coupling 2 is reinforced by large medially-directed flanges on the ceratobranchials 4 which support the lower pharyngeal jaws from below and lift them when the ceratobranchials 4 are lifted (Fig. 7). In the zooplanktivorous L. mucrochirus (Keast, 1977; Werner and Hall,


Fig. 6. Simplified and schematic representation of the mechanical model of pharyngeal biting in the Cichlidae by Galis (1992). The upper pharyngeal jaws are simplified here as one bone, because of the

very strong coupling between them in cichlids (Liem, 1973). Arrows indicate the directions of the forces exerted by muscles. Fp is the force that is exerted on the prey. (A) Lateral view and (R) caudal view of the pharyngeal jaw apparatus.

Fig. 7. Ventral view of the branchial basket of the molluscivorous centrarchid Leppomis gihhosus. The flanges on ceratobranchials 4 support the lower pharyngeal jaws. The m. transversus ventralis anterior which runs between left and right ceratobranchials 4 forms a muscle sling below the lower pharyngeal

jaws. The flanges, m. transversus ventralis and the ligaments between ceratobranchials 4 and the lower pharyngeal jaws all maintain a strong connection (coupling 2. Fig. 1).

650 Galia and Druckcr

Pig. 8. Schematic illustration of the position of the m. transvcrsus ventralis anterior in transverse view in perciforms. (A) Primitive position (centrarchids). contraction Icads to joined movement of lowet

pharyngeal jaw and ceratobranchials. 4 (B) specialized position (most cichlida), contraction lads to clcvation of the lower pharyngeal jaw towards ccratobranchials, 4.

1979) similar flanges are present but of smaller size and in the piscivorous M. salmoides (Werner, 1977) they are absent. Thus, there is a correlation between the development of the supporting flanges on the fourth ceratobranchials and the magnitude of the forces that are likely to be exerted in the vertical direction upon the lower pharyngeal jaws by the levator muscles (see results of muscle stimulation experiments for prey processing technique of M. ~salmoi~/~.s in which mainly horizon- tul forces are employed).

We expect that because of this coupling, contraction of m. levator externus 4 or m. levator posterior elevates not only the lateral limb of epibranchial 4 and ceratobranchial 4, but also the lower pharyngeal jaws (Figs. 1 and 3).

In the Cichlidae, by contrast, coupling 2 is replaced by a more flexible connection: there are only ligaments rostrally, that connect the lower pharyngeal jaw with the central axis (copulae communis) and with the most rostra1 part of the fourth gill arch (Anker, 1989). These connections permit movement of the lower pharyngeal jaw with respect to ceratobranchials 4 (Galis, 1992; Galis and Terlouw, unpublished data). In most cichlids m. transversus ventralis anterior has become bipartite, with fibers originating from both the left and right fourth ceratobranchials and inserting on the keel of the lower pharyngeal jaw (Stiassny, 1981). There are no flanges on the fourth ceratobranchials as in L. gihhosus and L. macroclzivus. The position of ceratobranchials 4 is dorsal to the lower pharyngeal jaw (Figs. 5 and 8b) and so contraction of the m. transversus ventralis anterior leads to elevation of the lower pharyngeal jaw with respect to the stationary fourth ceratobranchials (Galis and Terlouw, unpublished data).

In the Centrarchidae there is a strong connection between epibranchials 4 and the lower pharyngeal jaws (Fig. 1: coupling 3) which is both ligamentous and muscular (m. adductor branchialis 5, m. obliquus posterior; Wainwright, 1989). This direct connection further strengthens coupling 2 in linking the movements of the fourth epibranchials and the lower pharyngeal jaws.

Coupling 3 has been retained in cichlids (Fig. 4). Anker ( 1978, 1989) interprets the connective tissue linkage as part of the aponeurosis of m. levator externus 4 (see also Aerts, 1982; Claes and Aerts, 1984). M. levator cxternus 4 is thus still linked to epibranchial 4 via its myosept.

In cichlid species with the primitive insertion of m. levator posterior on the fourth epibranchials (e.g. 0. ni/otic~.s), coupling 3 is expected to transmit the force of this muscle to the lower pharyngeal jaw, as in centrarchids.

Pharyngeal biting models

Figures 3(a) and (b) present our mechanical model for pharyngeal biting in centrarchid fishes based on the upper pharyngcal jaw mechanism proposed by Wainwright ( 1989) and on the anatomical couplings described above. Figures 9 and lO( a) present schematic representations of the transmission of muscle force to prey.

The fourth epibranchial, linked both to the upper and lower pharyngeal jaws (Fig. l), plays a central role in our model: rotation of epibranchials 4 leads to coupled movements of both the upper and lower pharyngeal jaws that result in occlusion (Fig. 3). The predictions of the model are:

(1) Contraction of the m. levator posterior and tn. levator externus 4 leads to depression of the upper pharyngeal jaws as proposed by Wainwright (1989) (via coupling 1) and elevation of the lower pharyngeal jaws (via couplings 2 and 3).

(2) Contraction of the m. geniohyoideus causes elevation and protraction of the lower pharyngeal jaw (Fig. 3b), which implies that the m. geniohyoideus contributes to the force applied to the prey during biting.

(3) Contraction of the m. sternohyoideus leads to depression and retraction of the lower pharyngeal jaws (Fig. 2) which implies that the m. sternohyoideus contributes to the abduction of the pharyngeal jaws.

(4) Movements of the lower pharyngeal jaws interfere with those of the upper pharyngeal jaws and vice versa.

(5) The efficacy of force generation varies with the degree of rotation of epibranchial 4 (Fig. 9).

(6) The magnitude of the biting force varies with the size of the prey item (relative to the size of the pharyngeal jaws (Fig. 9).


Fig. 9. Schematic representation of force transmission in the proposed mechanical model of pharyngcal

biting in the Centrarchidae (caudal view). From top to bottom, the degree of rotation of epibranchial, 4 increases while the size of the prey decreases. The left and right sides of the tigure compare epibranchials of different shape. The degree of rotation of epibranchial 4, the shape of cpibranchial 4 and

the relative size of the prey all influence the force of the pharyngcal bitt. Length of arrows indicates relative magnitude of forces. FI is the force of the m. levator externus 4 and/or m. levator posterior. Fl* is the perpendicular component of FL Fp is the force exerted on the prey.


Fig. 10. Comparison of pharyngeal biting mechanics in centrarchids and cichlids (caudal view). A) Centrarchid model in which muscle force is fully transmitted to prey (cf. Fig. YD). B) Cichlid model (cf. Fig. 6). Length of arrows indicates relative magnitude of forces. Fl is the force of the m. levator externus 4 and/or m. levator posterior. Fp is the force exerted on the prey. Fr is the neurocranial reaction force. Note the doubling of the biting force (at a single point) that is the result of the union of the upper and

lower pharyngeal jaws in cichlids.

Fig. 11. Schematic representation of the mechanical model of pharyngeal biting in the Haemulidae by Wainwright (1989). Solid line arrows indicate the direction of the translation and rotation of the bones. Dashed line arrows indicate the direction of the forces exerted by muscles. Epibranchials 4 rotate around

the point where they are anchored to the lower pharyngeal jaws. An important difference between this model and the centrdrchid model of Fig. 3(a) is in the anchoring of the epibranchials. which in this model is supposed to be to the lower pharyngeal jaws. In centrarchids (and presumably haemulids) this is not possible: the lower pharyngeal jaws are elevated upon contraction of the pharyngeal levator muscles because of couplings (2 and 3, Fig. I) and can therefore not stabilize the central parts of

epibranchials 4 to function as a center of rotation (see text and Fig. 12).


Fig. 12. Photograph of the left lateral aspect of the orbital and pharyngcal regions of a cleared and

stained specimen of the ccntrarchid E~~rr~crr~rhu~s ~lorioms. (a) ovcrvicw. Scale bar = I mm. (b) detail 01 (a). A previously undescribcd ligamcntous structure (arrow) was found to lx suspcndcd mcdlally from the neurocranium, with one broad rostroventually-directed process and two slender dorsocaudally-directed processes. The structure is very thin, and presumably collagen Ibrtilicd wilh cartilage especially on the lateral sides (stained with Alcian blue). The central part of epibranchial 4 (indicated with a C) can

be moved up to this tentroof-like ligament. It is suggcstcd thaL blocking of epibranchials 4 by the ligament may be the mechanism lhat provides a fixed center of rotation for thcsc bones upon contraction of the m. Icvator posterior and m. Icvalor cx. 4 (see Fig. 3). Another function of this ligament may be

prelection of blood vessels from the action of the epibranchials 4. The slructure was also found in L. gihhmus, L. mc~c~rocl~rtus and M. strln~oic/~~.s.

Pharyngcal biting in pcrcilbrm fish 655

(7) The magnitude of the biting force varies with the shape of epibranchial 4 (Fig. 9).

Predictions I to 4 were tested in our experiments.

Wainwright ( 19X9) proposed in his model for haemulids that epibranchials 4 are anchored to the lower pharyngeal jaws (by the m. obliquus posterior and a ligament) and that this point of anchoring functions as the center of rotation (Fig. 11). The anchor point to the lower pharyngeal jaws, however, can only function as the center of rotation when the lower pharyngeal jaws are stationary. In out centrarchid model this is not the case. The strong connection between lower pharyngeal jaws and the fourth gill arches (epibranchials 4 + ccratobranchials 4) precludes this possibility, as can be seen in Figs. 1 and 3: epibranchial 4 and ceratobranchial 4 can not be lifted without also lifting the lower pharyngeal jaws (coupling 2 and 3). Thus, upon contraction of the pharyngeal levator muscles (Fig. 3), the lower pharyngeal jaws do not remain stationary but are pulled up togethrl with the lower limbs of epibranchials 4 (along with the ceratobranchials 4). Although the epibranchials 4 cannot be anchored as suggested by Wainwright (1989) for haemulids, our model requires that they are anchored to a structure outside the branchial apparatus such that their central part is stabilized and can function as a center of rotation when contraction of the m. levator posterior and the m. levator externus 4 causes elevation of the lateral limbs (Fig. 3).

We tested this implication of our model by looking for a structure that could anchor epibranchial 4. Study of centrarchid specimens cleared and stained for cartilage with Alcian blue revealed the presence of a ligamentous structure, rcsem- bling a tent-roof, rostra1 and dorsal to the medial part of both left and right epibranchial 4 (Fig. 12). The thin ligamentous structure, presumably of collagen, is fortified on the lateral sides with cartilage and is suspended from the neurocranium by three strands that arc also chondrified. In specimens cleared and stained for cartilage the ligament is highly conspicuous, but in unstained specimens it is difficult to find. Manual elevation of the lateral part of epibranchial 4 (comparable to the elevation caused by activity of the m. levatores externus 4 and m. levatores posterior) in freshly killed specimens of L. gihhosus showed that the central part of epibranchial 4 is blocked in this movement by the ligament. It seems possible that blocking by the ligament provides a fixed center of rotation for the bone (Fig. 3) upon contraction of m. levator externus 4 and m. levator posterior.

Cich lids

For cichlids we used the model of Galis ( 1992) (represented schematically in Fig. 6). In this model elevation of the lower pharyngeal jaw presses the prey against the


upper pharyngeal jaws. The upper pharyngeal jaws are supported by the neurocranium and therefore the neurocranial reaction force provides an important component of the biting force. The model assumes that a given muscle causes movement of either the upper pharyngeal jaw or of the lower one but not of both together. We tested this assumption and the following predictions of the model:

(1) Activity of the m. levator externus 4, the m. levator posterior and the m. transversus ventralis each result in elevation of the lower pharyngeal jaw.

(2) Activity of the m. pharyngocleithralis externus results in depression and retraction of the lower pharyngeal jaw.

(3) Activity of the m. pharyngocleithralis internus (Fig. 5) results in retraction of the lower pharyngeal jaw.

(4) Activity of the m. retractor dorsalis results in retraction of the upper pharyngeal jaws.

A comparison of’ the centrarchid and cichlid pharyngeal biting models

The biting mechanism of the centrarchid model is effective in a limited number of situations. The force of the m. m. levator posterior and levator externus 4 is fully transmitted to the prey only when it is applied in a direction perpendicular to the rotating lateral epibranchial 4 limb (Fl in Fig. 9). In all other orientations of the epibranchial 4 only the perpendicular component of the force is transmitted (Fl* in Fig. 9). Depending on the size of the prey items and the shape of epibranchial 4 the etfectiveness of the bite changes (Fig. 9). The relative length of the limbs of the epibranchial also influences bite effectiveness by affecting the moment (torque). In L. gihhosus both limbs are approximately of equal size so that no mechanical advantage is gained but in M. salmoides the lateral limbs are indeed longer than the medial limbs (Fig. l), thus conferring an advantage in leverage according to the model.

In the cichlid model there are no rotating bones and the force of the levator externus 4 and the levator posterior is always fully transmitted to the lower pharyngeal jaw (Galis, 1992; see also Fig. 6). Thus, a strong bite is possible for many different prey types. In cichlids the magnitude of the bite force is also affected by the fusion of the lower pharyngeal jaws and the strong bond between the upper pharyngeal jaws (Liem, 1973). In centrarchids forces on the lower and upper pharyngeal jaws are divided over the left and right jaws whereas in cichhds the forces are jointly exerted on single units, which doubles the maximum possible force (Fig. 10).

Results of muscle stimulation experiments

Centrarchids

In the experimentally determined actions of nine muscles of the branchial and hyoid skeletons are summarized in Table 1. The results are all in agreement with the


predictions of our model. Electrical stimulation of the muscles produced some movements of the pharyngeal jaws consistent with the findings of earlier work on non-labroid perciforms (Wainwright, 1989; Lauder, 1983a, b). However, structural couplings among the bony elements of the pharyngeal jaw apparatus, as described here, resulted in additional movements of the jaws not previously reported (bold face in Tab. 1).

Several muscles had an effect on both upper and lower pharyngeal jaws (Tab. 1). Not all muscles effecting movements of the lower pharyngeal jaws in the centrarchid species also caused movements of the upper pharyngeal jaws. Muscles further removed from the couplings between the upper and lower pharyngeal jaws (m. geniohyoideus, m. sternohyoideus, m. pharyngocleithralis externus) only produced movements of the lower pharyngeal jaws (Tab. 1).

In the piscivorous Micropterus salmoid~s the biting movements of the pharyngeal jaws are mainly protraction and retraction with very little depression and elevation. The upper pharyngeal jaws articulate with the posterior surface of epibranchials 4 via processes on the upper pharyngeal jaws that are positioned directly caudal to the fourth epibranchials (Fig. 1 b). The orientation of the m. levator posterior and the m. levator externus 4 is more horizontal than in the other centrarchid species (cf. Figs. 2a, b). Contraction of these muscles in M. sa1moide.s results, therefore. in pivoting of the fourth epibranchials in a more horizontal plane than in the other species, with the lateral limbs mainly protracting with little elevation and the medial limbs mainly retracting with little depression. The protraction of the lateral parts results in protraction of the fourth ceratobranchials and the connected lower pharyngeal jaws (Fig. 1 b). The processes on the upper pharyngeal jaws directly behind the medial parts of the fourth epibranchials ensure that retraction of the medial limbs causes retraction of the upper pharyngeal jaws. These processes are lacking in L. gihhosus and L. macrochirus, but also occur in the piscivorous L. gulosus, although they are smaller than in M. .sulmoida.

Cichlids

Stimulation of the branchial and hyoid muscles in cichlids caused movements of either the lower or the upper pharyngeal jaws, but never of both jaws together. This is in agreement with the independent mobility of the upper and lower pharyngeal jaws reported by Liem (1973) and Claes and DeVree (1991a, b) and with the assumption of the pharyngeal biting model of Galis ( 1992). The predictions of this model were confirmed. A result that was not predicted by the model was that stimulation of the m. sternohyoideus resulted in retraction and depression of the lower pharyngeal jaw.

In the three cichlid species studied, activation of the m. levator posterior and m. levator externus 4 resulted in elevation of the lower pharyngeal jaw with no noticeable movement of the upper pharyngeal jaws, or pivoting of the fourth epibranchials. As mentioned above, 0. niloticus has the primitive insertion of the m. levator posterior on the epibranchials 4. In spite of the fact that the epibranchials

65X Galis and Drucker

4 are suspended from the upper pharngeal jaws, stimulation of the m. levator posterior in this species did, thus, not affect movement of the upper pharyngeal jaws. The loss of coupling 1 (cf. Figs. 1 and 4) allows the epibranchials 4 to slide along the upper pharyngeal jaws, and presumably this sliding prevents the transmission of movement of the fourth epibranchials to the upper pharyngeal jaws.

Discussion

The anatomical couplings within the pharyngeal jaw apparatus of the Centrarchi- dae have important implications for the generalized perciform mechanism of pharyngeal biting. The fourth epibranchial, linked both to the upper and lower pharyngeal jaws, plays a central role in effecting pharyngeal jaw occlusion. For this role, the presence of a fulcrum structure (Fig. 12) that enables the rotation of epibranchial 4 is essential. Electrical stimulation of muscles inserting on epibranchial 4 confirms this central role of epibranchial 4, and supports the model of upper jaw depression proposed by Wainwright ( 1989) except for the anchoring of epibranchials 4 (see below). Wainwright demonstrated for the Haemulidae that contraction of the levator externus 4 and levator posterior causes medial rotation of epibranchial 4. Pivoting of this bone, and the transmission of force through the epibranchial-pharyngobranchial joint, depresses the upper pharyngeal tooth plate (Fig. I I). Stimulation of these muscles in the centrarchid specimens that we examined resulted in similar upper pharyngeal jaw depression (Tab. 1). This finding extends the applicability of the model of Wainwright (19X9) among percoids.

The mechanism of lower pharyngeal jaw elevation, however, differs from that proposed in the haemulid model. In the Haemulidae, the lower pharyngeal jaw is thought to be elevated by the m. protractor pectoralis, a muscle originating from the neurocranium and inserting tendinously on the toothed ceratobranchial 5. This insertion site is a derived character of the Haemulidae (Johnson, 1980; Wainwright, 1989). In this primitive condition, the protractor pectoralis has its insertion not on the branchial skeleton, but either directly on the pectoral girdle (cleithrum) or on the connective tissue closely adherent to the pectoral girdle (Greenwood and Lauder, 198 1). Thus, the model of lower pharyngeal jaw elevation presented by Wainwright (1989) cannot apply broadly to generalized perciforms. We propose in our model for the Centrarchidae, and for other taxa with the gcncralized perciform pharyngeal anatomy, that the couplings between epibranchial 4 and the lower pharyngeal jaw (Fig. l), and between ceratobranchial 4 and the lower pharyngeal jaw, transmit the force of contraction of levator extcrnus 4 and levator posterior to the lower pharyngeal jaws (Fig. 3). The result is lower jaw elevation. This prediction of the model was supported by our muscle stimulation experiments.


epihranchids) to both the depression of the upper phar~ngedjuw~s utd elevution of the her phuryngeal ,javs (Fig. 3).

Our conclusions on the biting mechanism in centrarchids are based in part on electrical stimulation of individual muscles. In reality muscles act together in a coordinated way. Nonetheless we think that the combined information gained from anatomical observations and manipulations of fresh specimens, and the confirmation of predicted effects of single muscles on the upper and lower pharyngeal jaws provides support for our mechanical model. Some additional support comes from electromyography experiments by Lauder ( 1983a, b) which show that the m. levator externus 4, m. levator posterior, m. geniohyoideus and m. pharyngocleithralis internus are all active during crushing of snails in L. gihhosus and L. rnicdophus. However, the activity of these muscles during biting has limited value for the interpretation of the model, because biting is usually a forceful activity. Such activities are characterized by the simultaneous activity of all involved muscles, antagonists and protagonists alike (Liem, 1978; Elshoud-Oldenhave and Osse, 1979; Sanderson, 1988; Galis et al., 1994) and it is thus not surprising that the motor pattern of mollusc crushing in centrarchids (Lauder, 1983a, b) is very similar to that of mollusc crushing in cichlids (Galis et al., 1994) despite the different pharyngeal biting mechanism.

Anchoring of epihranchial 4

Wainwright ( 1989) proposed for haemulids that the epibranchial 4 rotates around its point of anchoring to the lower pharyngeal jaw (Fig. 11). The mccha- nism of lower pharyngeal jaw elevation proposed in our centrarchid model and confirmed in our experiments implies that the rotating mechanism in centrarchids cannot proceed as proposed in the haemulid model (see pharyngeal biting models). Instead, the ligament that is suspended from the neurocranium in centrarchids rostra1 and dorsal to the medial limb of epibranchial 4 (Fig. 12) could function as a centre of rotation. The presence of cartilage in the ligament indicates that, indeed, the ligament is subjected to tension and compression (Merrilees and Flint, 1980). In cichlids there is no rotation of epibranchials 4 and the ligament is absent. We tentatively propose that this ligament serves to anchor the fourth epibranchials, in the absence of any other apparent structures that could function as fulcrum. However, a more detailed study of this structure and of the pivoting of epibranchial 4 needs to be undertaken.

Interestingly, in the Haemulidae dorsal rotation of the lateral limbs of epibranchial 4 by m. levator externus 4 and m. levator posterior caused elevation of the posterior part of the lower pharyngeal jaws (Wainwright, 1989; p. 237) as in centrarchids. This indicates that haemulids also possess the coupling between epibranchial 4 and lower pharyngeal jaw (Fig. I: coupling 3) and, thus, the important coupling between upper and lower pharyngeal jaws during biting. The presence of the coupling between the upper and lower pharyngeal jaws in both families underlines its importance in the generalized perciform biting mechanism.


Furthermore, the elevation of the lower pharyngeal jaws in haemulids in response to activity of the levator muscles implies that, as in centrarchids, the lower pharyngeal jaw cannot function as anchor for the fourth epibranchial as is assumed in the model of Wainwright (1989). There must thus be a different fulcrum structure.

Phnryt~ged biting mrchanism 61 cichlidy anu’ compparison with cot trarchids

The results of muscle stimulation experiments confirm previous observations that, unlike in centrarchids, movements of the upper and lower pharyngeal jaws are independent in the cichlids (Liem, 1973; Claes and De Vree, 199la and b; Galis, 1992, 1993a). Individually stimulated muscles caused movement of either the upper pharyngeal jaw or of the lower pharyngeal jaw, but never of both jaws simulta- neously (Tab. I),

In the labroid families Cichlidae and Labridae the power stroke of the pharyngeal bite is lower pharyngeal jaw elevation (Liem, 1973, 197X; Liem and Sanderson, 19X6; Claes and de Vree, 1991a and b; Galis and Terlouw, unpublished data). Prey is pressed against the upper pharyngeal jaws which are supported by the neurocranium (Figs. 6 and lob). Unlike generalized percoids, all labroids possess a facet on the upper pharyngeal jaws which articulates with a pharyngeal apophysis on the neurocranium (forming a diarthrosis, Fig. 5) suggesting that the upper pharyngeal jaws are pressed against the neurocranium by elevation of the lower pharyngeal jaws. This type of biting mechanism is considerably different from that of haemulids (Wainwright, 19X9) and centrarchids in which upper pharyngeal jaw depression plays an important role. Nevertheless, many of the pharyngeal muscles have retained similar functions in the two groups. The effects of stimulation of the muscles are very similar in centrarchids and cichlids (Tab. I), m particular those muscles that induce movements of the lower pharyngeal jaw.

Interestingly, the biting movements in the piscivorous centrarchid M. salmoides are antero-posteriorly oriented as in piscivorous cichlids (Witte and Barel, 1976; Liem, 197X). Despite very different biting mechanisms, horizontal scraping has thus independently evolved in both families for the processing of fish prey. Similarly a mainly vertical biting force is employed in both families when hard molluscs are crushed (Witte and Barel, 1976; Claes and De Vree, 1991a; Galis and Terlouw, unpublished data).

The biting mechanism of the centrarchids can be effective only in a limited number of situations, since the magnitude of the biting force varies with the angle of rotation (Fig. 9). Depending on the relative size of prey items and epibranchial 4 and on the shape of epibranchial4, the force of the bite changes. This should limit the range of prey items that can be processed by centrarchids. In cichlids epibranchial rotation does not play a role in biting and the force of m. levator externus 4 and m. levator posterior is fully transmitted to the lower pharyngeal jaw regardless of prey size (Figs. 6 and lob), thus allowing a diet of many different prey types. The uncoupling of the upper and lower pharyngeal jaws in cichlids further allows the generation of biting forces in many different directions (Galis, 1992). In


addition, the force that can be applied to the prey is increased by the fusion of the lower pharyngeal jaws and the strong bond between the upper pharyngeal jaws in cichlids (Liem, 1973). In centrarchids forces on the upper pharyngeal jaws and lower pharyngeal jaws are divided over the left and right jaws (although the lower pharyngeal jaws are sometimes partly fused, e.g. in M. salmoicles) whereas in cichlids the forces are jointly applied on single mechanical units, doubling the maximum possible force that can be exerted on the prey (Fig. 10). The pressure that is applied locally depends on the area of contact which will vary tremendously among species, yet the possibility of exerting considerably greater forces has been created by the union of the left and right pharyngeal jaws.

The coupling of the upper and lower pharyngeal jaws in centrarchids seriously constrains the number of possible biting movements and, thus, the number of activity patterns of the pharyngeal jaw muscles. It is not to be expected for example that the piscivorous M. .su/~zoi~~s with its horizontally directed bite will be able to make vertical biting movements nor that the mollusc crushing L. gihhosus will be able to make strongly horizontal biting movements. Indeed the variety of muscle activity patterns in pharyngeal biting seems to be lesser in centrarchids (Lauder, 1983; Wainwright and Lauder, 1986) than in cichlids (Galis et al., 1994; see also Galis, 1993a). However more research is necessary to confirm this expectation.

Labroids are thought to be most closely related to the Percoidei (Kaufman and Liem, 1982; Stiassny and Jensen, 1987). All examined representatives of both generalized percoid families and percoid families considered to be close relatives of the derived Labroidei display the epibranchial rotation typical of haemulids and centrarchids (Wainwright, pers. comm.; generalized percoid families examined: Lutjanidae, Serranidae, Centrarchidae, Perchichthyidae, Centropomidae; Percoid families closely related to labroids : Lethrinidae, Haemulidae and Kyphosidae). Wainwright (pers. comm.) even observed this pharyngeal biting mechanism in the primitive Actinopterygians Amiu and Lepisosteus. As discussed above, the pharyngeal jaw apparatus of haemulids shares another important characteristic of the centrarchid pharyngeal jaw apparatus: the coupling between upper and lower pharyngeal jaws (Fig. 1, coupling 3). Representatives of the Percidae and Lethrinidae were also found to possess this structural connection (Galis, unpublished data) which in combination with epibranchial rotation provides for a coupling between the upper and lower pharyngeal jaw during biting. In the following transformation scenario it is, therefore, assumed that the labroid pharyngeal jaw apparatus is derived from a pharyngeal jaw apparatus that is, in essence, like that of the generalized perciform family Centrarchidae (i.e. characterized by epibranchial rotation and coupling between upper and lower pharyngeal jaws). However, we stress that this does not represent evidence that labroids are more closely related to centrarchids than to any other percoid family. The transformation scenario that we present is, following Galis ( 1996) based on mechanical arguments


Table 2. Proposed sequence of events in the evolution of the decoupled labroid pharyngeal jaw apparatus.

Anatomical modification Proposed functional consequences

(I) Decoupling of EB4 from UPJ (in Independent movement of UPJ and association with the development of a LPJ, labroid type pharyngeal bite

diarthrosis between neurocranium possible; incrcascd

and uppcr pharyngeal jaws) biting force

(2) Shift of insertion of LE4 from EB4

to LPJ More forceful elevation of LPJ

(3) Decoupling of LPJ from CB4

(4) Origin of m.tr. ventr. ant. on CB4

and insertion on LPJ (in association with a more dorsal position of cpihranchials 4 and with the

development of a median keel on LPJ as site of insertion of the muscle)

Increased manoeuvrability of LPJ

Increased manocuvrability and more

forceful elevation of LPJ

(5) *Twisting of fibers of m. retractor dorsalis*

Increased biting force; increased manocuvrability of UPJ

(6) *Fusion of left and right LPJ**, and

incrcascd union of left and right UPJ

Increased biting force; incrcascd control

of LPJ and UPJ

* Order of events not clear in transformation sequence. ** The degree of fusion increased throughout the transformation For abbreviations see Table I.

and not upon a known phylogenetic hypothesis. It is the scenario that is most feasible biomechanically, similar in approach to that of Lombard and Wake ( 1986) on the evolution of the tongue in plethodontid salamanders. In the evolution of the cichlid pharyngeal jaw apparatus, two structural couplings that are primitive for perciforms have been modified. First, the coupling between epibranchial 4 and the upper pharyngeal jaws (Fig. 1: coupling 1) has become flexible and changed such that force exerted on epibranchial 4 by the m.m. levatores posterior and externus 4 does not result in concomitant movement of the upper pharyngeal jaw. Only the lower pharyngeal jaw is moved by these muscles. Second, the coupling between ceratobranchial 4 and the lower pharyngeal jaw (Fig. I : coupling 2) has become flexible, permitting extensive independent movement of the bones. The strong ligamentous connection between epibranchial 4 and the lower pharyngeal jaw present in centrarchids (Fig. 1: coupling 3) has been retained in cichlids. Based upon our results we propose the following sequence of anatomical changes that arose in the evolution of the decoupled labroid pharyngeal jaw apparatus (summarized in Table 2).

Step 1. We propose that the key innovation in the development of the labroid pharyngeal bite was the decoupling of epibranchial 4 movements from those of the upper pharyngeal jaws (loss of coupling 1) and its crucial consequence, the uncoupling of upper and lower pharyngeal jaw movements. The decoupling stems


Fig. 13. Proposed mechanical model of pharyngeal biting in a perciform fish that exhibits decoupling of

the upper pharyngeal jaws from the epibranchials 4. the first step in the transformation from the generalized pcrcoid (e.g. centrachid) to derived perciform (e.g. cichlid) pharyngcal jaw apparatus. The levator muscles retain their primitive insertion of epibranchials, 4. Caudal view. solid and dashed arrows indicate the direction of the forces. Grey arrows indicate the sliding movements of epibranchials 4 upon

contraction of the pharyngeal levator muscles. The epibranchiak 4 do not rotate, as they do in ccntrarchids (Fig. 3A), and instead are elcvatcd. Since movement of the epibranchials, 4 do not induce movement of the upper pharyngeal jaw, the elements are decoupled.

from relatively minor structural changes in the primitive percoid pharyngeal jaw apparatus: a change in the articulation between the fourth epibranchials and upper pharyngeal jaws such that sliding of the former along the latter becomes possible (Figs. 4 and 13). The direct contacts between the upper pharyngeal jaws and neurocranium that are a consequence of the decoupling (Fig. 13) require a cartilaginous cushion, such as occurs in the generalized percoids Buds and Pvis- tolepis (Liem, 1973).

The results of muscle stimulation experiments on the cichlid Oveochromis niloticus suggest that decoupling of upper and lower pharyngeal jaw movements was the first evolutionary step in the transformation from a coupled to a decoupled pharyngeal biting mechanism. The pharyngeal jaw apparatus of 0. niloticus has undergone an incomplete transformation from the generalized condition found in centrarchids to the derived condition of many other labroids. The coupling between the upper pharyngeal jaw and epibranchial 4 (Fig. 1: coupling 1) has become as flexible as in the other cichlids examined in this study. However, the m. levator posterior retains its primitive insertion on epibranchial 4. Stimulation of the m. levator posterior elevates and protracts the lower pharyngeal jaw through the coupling between epibranchial 4 and the lower pharyngeal jaw (Fig. 4: coupling 3), but causes no movement of the upper pharyngeal jaw due to the derived uncoupling of epibranchial 4 from the upper pharyngeal jaw. The result is a labroid type of pharyngeal bite, in which the lower pharyngeal jaw is elevated against the upper pharyngeal jaw as the latter is stabilized by the neurocranium (Fig. 13). Therefore it is possible to have a labroid-type bite with both m. levator posterior and m. levator externus 4 inserting in the primitive perciform location on epibranchial 4 and both couplings 2 and 3 still present. The immediate advantage of this decoupling (loss of coupling 1) is the reduction in muscular force required in biting: the force transmission to the lower pharyngeal jaw improves the neurocranial


reaction force provides most of the downward-directed force of the upper pharyngeal jaw (Galis, 1992).

The shift in insertion of the m. levator externus 4 from epibranchial 4 to the lower pharyngeal jaw, which was proposed as the key structural innovation of the labroid pharyngeal jaw apparatus by Liem (1973) seems most advantagenous qfter the decoupling of movements of the upper and lower pharyngeal jaws. The shift in insertion unarguably contributes to the mechanical flexibility of the lower pharyngeal jaw (Liem, 1973, 1978, 1979; Claes and DeVree, 1991a and b), but can only do so after decoupling of the upper and lower pharyngeal jaw movements. Before this decoupling, the shift of insertion of the m. levator externus 4 would diminish the effect of upper pharyngeal jaw depression because it would reduce the force transmission of this muscle to the upper pharyngeal jaw. In our experiments with centrarchids, only the muscles that were directly attached to the centrally positioned epibranchial 4 produced powerful movements of both upper and lower pharyngeal jaws. The absence of a muscular sling in most species of the labroid family Pomacentridae and the presence of a diarthrosis between the neurocranium and upper pharyngeal jaws throughout the Pomacentridae ( Liem and Greenwood, 198 1; Kaufman and Liem, 1982; Stiassny and Jensen, 1987) gives support to the hypothesis that decoupling of upper and lower pharyngeal jaw movements occurred before the shift in insertion of m. levator externus 4 (Liem and Greenwood, 1981; Kaufman and Liem, 1982).

Step 2. After the decoupling of lower and upper pharyngeal jaw movements, the shift of insertion of m. levator externus 4 from epibranchial 4 to the lower pharyngeal jaw provides a more force-effective bite. When m. levator externus 4 has its primitive insertion on epibranchial 4 it elevates the lower pharyngeal jaw indirectly (Fig. 13). The connection between epibranchial 4 and the lower pharyngeal jaw is formed by a ligament and by two muscles, m. adductor branchialis 5 and m. obliquus posterior (coupling 3). Forceful contraction of the m. levator externus 4 requires contraction of these muscles in order to maintain the connection. After the shift of m. levator externus 4 to the lower pharyngeal jaw only the force of m. levator posterior (which still inserts on the epibranchials 4) needs to be counter- acted to maintain the connection.

Step 3. Once m. levator externus 4 gained its insertion on the lower pharyngeal jaw, decoupling of ceratobranchial 4 from the lower pharyngeal jaw became highly advantageous because it allowed greater manoeuvrability and range of motion of the lower pharyngeal jaw. Presumably, the disadvantage of this decoupling is reduced stability of the lower pharyngeal jaw during biting. Decoupling of the lower pharyngeal jaw and ceratobranchial 4 prior to the shift in insertion of m. levator externus 4 does not seem to confer a functional advantage because no increase in manoeuvrability of the lower pharyngeal jaw would be gained. Hence we propose that the shift in insertion of the levator externus 4 from epibranchial 4 to the lower pharyngeal jaw preceded the decoupling of the lower pharyngeal jaw from ceratobranchial 4 (Tab. 2).

Step 4. In the primitive perciform pharyngeal jaw apparatus the m. transversus ventralis anterior interconnects the margins of the left and right ceratobranchials 4


while holding the lower pharyngeal jaws in a muscular sling (Figs. 7 and g(a)) (Winterbottom, 1974; Stiassny, 1982; Wainwright, 1989). Contraction of this muscle adducts ceratobranchials 4, thereby adducting the lower pharyngeal jaws (Wainwright, 1989). When inactive, the muscle probably functions as a ligament further restricting independent movements of the lower pharyngeal jaws and ceratobranchials 4. In its derived condition, the m. transversus ventralis anterior does not run as a muscular sling below the lower pharyngeal jaws connecting left and right ceratobranchials 4, but instead extends from each ceratobranchial 4 to the lower pharyngeal jaw itself (Figs. 5 and 8b) (Stiassny, 1982; Stiassny, 1990; see also Anker, 1978). Associated with the derived position of the m. transversus ventralis anterior is the possession of a well-developed median keel on the ventral face of the lower pharyngeal jaw, which serves as site of attachment for the muscle (Fig. 8b; Stiassny and Jensen, 1987). According to Aerts (unpublished data) the shift in insertion of the m. transversus ventralis is induced by the development of this keel. The position of the fourth ceratobranchials at the place of origin of the m. transversus ventralis is dorsal to the lower pharyngeal jaw (Figs. 5 and 8b). Therefore, in cichlids contraction of the m. transversus ventralis anterior results in elevation of the lower pharyngeal jaw (Tab. 1) towards the fourth ceratobranchials. The derived insertion on the lower pharyngeal jaw increases the manoeuvrability of the lower pharyngeal jaw. Thus, decoupling of the lower pharyngeal jaw from ceratobranchial 4 (Tab. 2, step 3), followed by the more dorsal placement of the fourth ceratobranchials and the shift in insertion site of the transversus ventralis anterior on the keel of the lower pharyngeal jaw (step 4), led to a highly manoeuvrable lower pharyngeal jaw in labroid perciforms. The derived insertion of the m. transversus ventralis anterior is mosaically distributed in the Labroidei (Stiassny, 1982 and 1990) suggesting that it is a character which emerged several times independently.

Step 5. Another anatomical change increasing the flexibility of the cichlid pharyngeal jaw apparatus is a twist in the fibers of the retractor dorsalis which results in several lines of action for this one muscle (Tab. 2; Galis, 1992). The presence of these ditferent lines of action greatly increases the effectiveness of the bite in most situations (Tab. 2; Galis, 1992). It is not clear when in the transformation sequence the retractor dorsalis was modified but the change was undoubt- edly advantageous after decoupling of upper and lower pharyngeal jaw movements (step I).

Step 6. Partial fusion of the lower pharyngeal jaws occurs in some generalized perciforms. The fusion is markedly increased in cichlids and more so in other labroids (Kaufman and Liem, 1982; Stiassny and Jensen, 1987). In cichlids the bond between the upper pharyngeal jaws is strong allowing them to function as a single bone (Liem, 1973). The advantage of the union of upper and lower pharyngeal jaws is that the maximum force generation is doubled (Fig. 10). An important additional advantage is improved control over prey manipulation because the number of muscles controlling the movement of the fused jaw is effectively doubled. A disadvantage is a loss in degrees of freedom because one instead of two pharyngeal jaws can be moved.


The results of the multi-step anatomical transformation of the generalized perciform pharyngeal jaw apparatus (Tab. 2) was the structurally decoupled, functionally versatile pharyngeal jaw apparatus of the Labroidei. The derived apparatus is capable of processing a wide range of prey types using a broad repertoire of different feeding modes (Fryer and Iles, 1972; Greenwood, 1974; Liem, 1973, 1978; Liem and Greenwood, 1981; Barel, 1983; Claes and DeVree, 1991 a, b; Witte et al., 1992; Galis et al., 1994).

Vcrmeij ( 1974) proposed that an increase in the number of independent structural elements in a musculoskeletal system increases the number of degrees of freedom and, hence, allows a greater number of possible mechanical solutions for functional problems. Analogously, an increase in the number of structural elements in general will lead to a greater potential diversity of form and function. With an increase in the number of solutions for functional problems comes the possibility of greater mechanical efficiency and effectiveness of resource exploitation (Vermeij, 1974). He hypothesized that more versatile taxa or body plans tend to replace less potentially versatile taxa in the course of time (see also Schaeffer and Rosen, 1961). Lauder ( 1981) and Lauder and Liem ( 1989) argued further that taxa exhibiting a large number of structural decouplings are expected to be more versatile and hence more speciose than taxa with fewer decouplings (see also Galis. 1996).

In this study, cichlid fishes were selected as representative perciforms with a decoupled pharyngeal jaw apparatus. Correlated with this anatomical specialization in cichlids is the occupation of a vast array of feeding niches and a great species-richness. The occupation of almost any feeding niche available in freshwater is seen as a conspicuous characteristic of cichlid speciation by most biologists studying cichlids (e.g. Fryer and Iles, 1972; Greenwood, 1974, 1984; Liem, 1980; Lewis, 1981; Kaufman and Liem, 1982; Barel, 1983; McKaye and Marsh, 1983; Ribbink et al., 1983; Sage et al., 1984; Witte, 1984; Dominey, 1984; Capron de Caprona and Fritzsch, 1984; Lowe-McConnell, 1987; Stiassny and Jensen, 1987; Yamaoka, 1991; Meyer, 1991). Liem (1973, 1978, 1979, 1980) proposed the versatility of the pharyngeal jaw apparatus as a key factor in the explosive radiation of cichlids. Indeed, on the one hand this versalitity allows cichlids to eat a large number of different prey types (Liem, 1978, 1979; Galis 1993; Galis et al., 1994) so that they can survive in adverse and changing environments. On the other hand, the bauplan is versatile in that only small evolutionary changes arc necessary to make it suitable for the processing of novel prey items (Liem, 1973; Greenwood, 1973).

The suborder Labroidei (which includes the Cichlidae) is as a whole unusually spcciose and trophically diverse (Jensen, 1990) and is united by a suite of specialized features of the pharyngeal jaw apparatus including the articulation between the upper pharyngeal jaws and the neurocranium (Kaufman and Liem, 1982; Stiassny and Jensen, 1987). This suggests that the versatility of the decoupled pharyngeal jaw apparatus may be fundamental to the evolutionary success of not only cichlids but


of the entire suborder. The importance of the changes in the pharyngeal jaw apparatus in the evolutionary development from generalized perciforms to labroids is emphasized by the fact that 7 out of 8 characters distinguishing the labroids from other perciforms are characters of the pharyngeal jaw apparatus (Stiassny and Jensen, 1987). Further support for this hypothesis comes from a comparative analysis of Jensen (1990) which shows not only that labroids (appr. 1800 species) are more speciose than any of their possible sister groups, but also that Exo- coetoidea (Beloniformes), a group of fishes which have many of the specialisations of the pharyngeal jaw apparatus that distinguish the labroids from other perciforms, are more speciose than their sister group within the Beloniformes, the Scomberesocoidea ( 135 + species versus 36).

Obviously, other mechanisms, e.g. behavioural mechanisms, have played a role in the evolution and speciation of cichlids and other labroids. We believe, however, that the transformation from a primitive and coupled perciform pharyngeal jaw apparatus to the derived and decoupled form in labroids is a key evolutionary innovation that facilitated speciation and allowed a remarkable diversification of form and function.

Acknowledgements

We thank Jacques van Alphen. Coen van den Bergh. Jos van den Roogaerd. Beth Brainerd. Rcn6 Hcngchnolen, Rob Iloutman, Les Kaufman, Karcl Liem. Ilans MetT, Mees Mullcr. Jan Scvcnstcr, Andy

Sih, Peter Wainwright and Ernie Wu for constructive discussions. Mccs Mullcr is especially thanked for his help with the biomechanical analysis. Gerrit Anker kindly supplied us with the originals of his reconstruction drawings. Pctcr Aerts kindly provided unpublished Information on the ontogeny of the m.

transvcrsus ventralis and Peter Wainwright on pcrciform biting. We thank Beth Brainerd. Alex Carbo. Karstcn Hartcl, Pat Hernandey, Kees Hofker, Donna Ncmcth. Martien van Ooijen and Aric Tcrlouh for assistance with the experiments and Pctcr Aerta, Gerrit Anker. Paul Brakefield. Allen Herre. Jeff Jensen, Karel Licm. John Long. Mees Muller, Nand Sihbing. Andy Sih. Peter Wainmright and Frans

Wittc for comments on carlicr versions of this article. Andrie ‘t Hooft took the photographs. This rcsearch was supported by a Fulhright scholarship. 2, travel grant of the Netherlands Organisa-

tion for Scientific Rcscarch (NWO) and a grant of the J. L. Dobbcrkc I:oundation to Frietson Calis and NSF grant BSR 88-18014 to Karel F. Liem.

References

Aerts, P. 1982. Development of the musculus levator cxtcrnus IV and the musculus obliquus posterior in /luphhron/i.c &~~u’(III.Y (Trewavas, 1993 Tclcostci: Cichlidue). A discussion on the shift hypothe-

sis. J. Morph. 173: 225 235. Anker. G. Ch. 1978. The morphology of the head-muscles of a gcncralizcd Haplochromis hpeciei: H.

c/~~,~un,s Trewavas 1933 (Pisces, Cichlidae). Ncth. J. 2001. 2X: 234 271. Ankcr, G. Ch. 1989. The morphology of joints and ligaments in the head of a generalized I/tr/,io(,h~o,?li.v

species: II. E/~,~rm.s Trewavas 1933 (Teleostei, Cichlidae). III. The hyold and the branchiostegal apparatus. the branchial apparatus and the shouldcl- girdle apparatus. Neth. J. Zool. 39: I 40.

Harel, C. D. N. 1983. Towards a constructional morphology of cichlid fishes (Teleoqtei. Perciformcs). Ncth. J. Zool. 33: 357 424.


Bare], C. D. N., F. Witte and M. J. P. van Oijen. 1976. The shape of the skeletal elements in the head

of a generalized Huplochromis spccics: If. c+~cms Trewavas 1933 (Pisces, Cichlidae). Neth. J. Zool. 26: 163-265.

Capron de Caprona, M. D. and B. Fritzsch. 1984. Interspecific fertile hybrids of haplochrominc Cichlidae and their possible importance of speciation. Neth. J. Zoo]. 34: 503--538.

Claes, G. and F. De Vrcc. 199la. Kinematics of the pharyngeal jaws during feeding in Owochrornis

rdotiws (Pisces, Perciformes). J. Morph. 208: 227 245. Claes, G. and F. De Vrec. 1991b. Cineradiographic analysis of the pharyngeal jaw movements during

feeding in Hqdochrow~i~s hurtonr (Giinther, 1893) (Pisces, Cichlidne). Belg. J. Zool. 121: 227 234. (‘lacys, H. and P. Acrts. 1984. Note on the compound lowcr pharyngeal jaw operators in As/rrfotr/q>iu

c,/cgtm.~ (Trewavas), 1933 (Teleostei: Cichlidac). Neth. J. Zoo]. 34: 210 -214. Dominey, W. J. 1984. Erects of sexual selection and life history on speciation: Species ltocks of African

cichlids and Ilawaiian Drosophila. Iw A. A. Echelle and I. Kornfield (Eds.), Evolution of Fish

Species Flocks. University of Maine and Orono Press, Orono, Maine. Druckcr, E. G. and J. S. Jensen. 1991. Functional analysis of a specialized prey processing behavior:

winnowing by surfperches (Teleostei: Embiotocidae). J. Morph. 210: 267 287. Elshoud-Odenhave, M. J. W. and J. W. M. Ossc. 1979. Prey capture in the pike-perch. Stizostctlio~~

/~rc,iqx,rc,cr (Teleostei. Percidae): A structural and functional analysis. Zoomorphologie 93: I 32.

Fryer. G. and T. D. Iles. 1972. The Cichlid Fishes of the Great Lakes of Africa. Oliver and Boyd. Edinburgh.

Galis, F. 1992. A model for biting in the pharyngeal jaws of a cichlid fish: Hqdochwn~i.~ piwtrtrr.s. J.

Thcor. Biol. 155: 343-368. Galis. F. 1993a. Interactions between the pharyngcal jaw apparatus, feeding behaviour and ontogeny in

the cichlid fish, ffqolochron~i.s piwoms. A study of morphological constraints in evolutionary

ecology. J. Exp. Zoo]. 267: 137p 154. Galia. F. 1993b. Morphological constraints on behaviour through ontogeny. The importance of

dcvclopmcntal constraints. Mar. Behav. Physiol. 23: 119~- 135. Galis. F. 1996. The application of functional morphology to evolutionary studies. Trends Ecol. Evol.

l](3): 124 129. Galls, I-. A. Tcrlouw and J. W. M. Osse. 1994. The relation betwcen morphology and behaviour during

ontogcnctic and evolutionary changes. J. Fish Biol. 45 (suppl. A): I3 26.

Greenwood, P. H. 1965. Environmental effects on the pharyngcal mill of a cichlid fish, A.v/tr/orcoc,lzronli.c trllrrtrrrtli. and their taxonomic implications. Proc. Linn. Sot. Lond. 176: 1 10.

(;reenwood, P. H. 1973. Morphology, endemism and speciation in African cichlid fishes. Verh. dt. zoo1. Ges. Mainr. 66: I 15- 124.

Greenwood, P. H. 1974. The cichlid fishes of Lake Victoria East Africa: the biology and evolution of a species Ilock. Bull. Hr. Mus. Nat. Hist. (Zool.) Suppl. 6: I 134.

Greenwood, P. H. 1984. African cichlids and evolutionary theories. In A. A. Echelle and I. Kornfield

(Eds.), Evolution of Fish Species I:locks. University of Maine and Orono Press, Orono, Maine. Greenwood, P. H. and G. V. Lauder. 1981. The protractor pectoralis muscle and the classification of

teleost fishcs. Bull. Br. Mus. Nat. Hist. (Zool.) 41(4): 213-234. Greenwood, P. H.. D. E. Rosen, S. II. Weitrman and G. S. Myers. 1966. Phyletic studies of teleo-

stcan fishes, with a provisional classification of living forms. Bull. Am. Mus. Nat. Hist. 131(4):

341 455. Hoogcrhoud, R. J. C. and C. D. N. Bare]. 1978. Integrated morphological adaptations in piscivorous

and mollusc-crushing Hrrplochmnis spccics, pp. 5 2 56. In J. W. M. Ossc (Ed.). Proceedings of the Zociac Symposium on Adaptation. Pudoc, Wageningen, The Netherlands.

Jensen, J. S. 1990. Plausibility and t&ability: Asaesbing the consequences of evolutionary innovation. 1?1 M. H. Nitccki (Ed.), Evolutionary Innovations. The University of Chicago Press, Chicago and

London. Johnson, G. D. 1980. The limits and relationships of the Lutjanidae and associated families. Bull.

Scripps Inst. Oceanogr. 24: I-114.

Pharyngcal biting in perciform fishes 669

Kaufman, L. and K. F. Licm. 1982. Fishes of the suborder labroidci (Pisces: Perciformes): Phylogeny. ecology. and evolutionary significance. Brcviora (Museum of Comparative Zoology, Ilarvard

University) 472: I 19. Kcast, A. 1977. Mechanisms minimizing intraspecific competion in vertebrates. with a quantitative study

of the contrasting strategies of two centrarchid fishes. ,4~?7hl0p/ir~:~ r~~pr.s~ri.~ and Lqwwi.~ nxlcrochirus. Evol. Biol. IO: 333 335.

Keast, A. 197X. Feeding interrelations between age-groups of pumpkinseed (L~,,x~rrrir ~ihho.su.c) and comparisons with bluegill (L. !nrrcroc/rirus). J. Fish. Res. Board Can. 35: I2 27.

Lauder, G. V. 1981. Form and function: struclural analy& in evolutionary morphology. Paleobiologq.

I: 430-442. Lauder, G. V. 19X3:1. Functional and morphological bases of trophic spcclali/ation in sunlishes

(Tclcostei, Ccntrarchidae). J. Morphol. 178: I 21, Lauder, G. V. 1983b. Neuromuscular patterns and the origin of trophic specialization in fishes. Science

219: 123.5 1237. Lauder, G. V. and K. F. Lie,. 1983. The evolution and interrcl;ltionshipa of the actinoprygian f&he.

Bull. Mus. Camp. Zool. Harvard University 150: 95 197.

Lauder. G. V. and K. 1:. Liem. 1989. The role of historical factors in the evolution of complex organismal functions. /n Wake. D. B. and G. Roth (I&.). Complex Organismal Functions: Integration and Evolution in Vertebrates. Dahlcm Conference. West Berlin. 19X8. John Wiley and Sons Ltd, New York.

Lewis, D. C. S. I98 I. Preliminary compat-isons between the ecology of the haplochromine cichlid fishes of Lake Victoria and Lake Malawi. Neth. .I. Zool. 31: 746 761.

Licm. K. F. 1970. Comparative functional anatomy of the Nandldae (Pisces: Telcosti). Ficldiana (Zoology) 56. I 166.

Licm. K. F. 1973. Evolutionary strategies and morphological innovatlons: (‘ichlid pharyngeal jaws. Syst. Zool. 22: 42s 441.

Liem, K. F. 197X. Modulatory multiplicity in the functional repertoire of the feeding mechanisms in cichlid fishes. J. Morph. 158: 323 360.

Liem, K. F. 1979. Modulatory multiplicity in the feeding mechanism in cichhd fishes. as examplified by the invert&rate pickers of Lake Tanganyika. J. Zool., Lond. 189: 93-- 12.5.

Licm, K. F. 1980. Adaptive significance of intra- and interspecific ditfercnces in the feedmg repertoire of

cichlid fishes. Amer. Zool. 20: 295-314. Liem, K. F. and P. H. Greenwood. 1981. A functional approach to the phylogeny of the pharyngognath

teleosts. Amer. Zool. IS: X3 IO I Liem, K. F. and S. L. Sanderson. 1986. The pharyngeal jade apparatus of labrid fishes: A functional

morphological perspective. J. Morph. 1X7: l43- 15X. Lombard. R. E. and D. B. Wake. 1986. Tongue evolution In the lungless salamanders family

Pletodontidac IV, Phylogeny of plethodontid salamanders and the evolution of feeding dynamics.

Syst. Zool. 35: 532 551. Lowe-McConnell, R. H. 19X7. Ecological studies in Tropical Fish Communities. CambrIdge Univ. Press,

Cambridge. McKaye, K. R. and A. Marsh 1983. Food switching by two specialircd algae-scraping cichhd fishes in

Lake malawi, Africa. Oecologia (Berlin) 56: 245 24X. Merrilecs, M. J. and M. H. Flint. 1980. Ultrastructural study of tension and pressure zones in a rabbit

flexor tendon. Amer. J. Anat. 157: X7 106. Meyer, A. 1991. Trophic polymorphisms in cichlid fish: Do they represent intermediate stcph dul-lng

sympatric speciation and explain their rapid adaptive radiation? I/I J. H. Schrocdcr (Ed.). New Trends in Ichthyology. Paul Party, Berlin.

Ribbink, A. J., B. A. Marsh, A. C. Marsh, A. C. Ribbink and B. J. Sharp. 1983. A preliminary survey

of the cichlid fishes of rocky habitats in Lake Malawi. S. Afr. J. Zool. IX: I 180. Sage, R. D., P. V. Loisellc, P. Basasibwaki and A. C. Wilson. 19X4. Molecular versus morphological

change among cichlid fishcs of Lake Victoria. h A. A. Echelle and 1. Kornfirld (Eds.). Evolution of Fish Species Flocks. University of Maine and Orono Press, Orono. Maine.

670 Galis and Drucket

Sanderson, S. L. lY88. Variation in neuromuscular activity during prey capture by trophic specialists and generalists (Pisces: Labridac) Brain Behav. Evol. 32: 257-268.

Schaeffcr, B and D. E. Rosen. 1961. Major adaptive lcvcls in the evolution of the actinopterygian

feeding mechanism. Amer. Zool. I: I87 204. Stiassny, M. L. J. 1981. The phyletic status of the family Cichlidac (P&es. Perciformcs): A comparative

anatomical invcatigation. Neth. J. Zool. 31: 275 314.

Stiassny, M. L. J. 1982. The relationships of the neotropical genus Cichla (Perciformcs, Cichlidae): a phyletic analysis including some functional considerations. J. Zool. Lond. 107: 427 -453.

Stiasuny. M. L. J. 1990. ~~k~~/zrorni,r, relationships and the phylogcnctic status of the African Cichlidae. Am. MUL. Nat. Hist. NY 2993, I4 pp.

Stlaasny. M. L. J. and J. S. Jcnscn. 1987. Labroid intrarclationships revisited: Morphological complexity, key Innovations. and the study of comparative diversity. Bull. Mua. Comp. Zool. 151: 269 319.

Vandcwalle, P.. M. Havard, G. Clacs and F. De Vree, 1992. Mouvements dcs machoircs pharyngienncs pendant la prise de nourriturc chcr le Serranus scribe (Linnk, 1758) (Pisces, Scrranidae). Can. J. Zool. 70: 145 160.

Vandcwnllc, P., A. Iluysseune, P. Acrts and W. Verraes. 1994. The pharyngeal apparatus in teleost feeding. Adv. Comp. Environ. Physiol. 18: SY 92.

Vermeij. G 1974. Adaptation, versatility, and evolution. Syst. Zool. 22: 466 477. Wainwright. P. C. 1987. Biomechanical limits to ecological performance: mollusc-crushing by the

Caribbean hogfish. Ltrc~/zno/trinzu.s nztrxin~lts (Labridac). J. ZooI., Lond. 213: 2X3-297.

Wainwright. P. C. 1988. Morphology and ecology: functional basis of feeding constraints in caribbean labrid fishes. Ecology 69: 635 64.5.

Wninwright. P. C. 1989. Functional morphology of the pharyngeal jaw apparatus in perciform fishes: An experimental analysis of the Haemulidae. J. Morph. 200: 231 245.

Walnwright, P. C. and G. V. Lauder. 1986. Feeding biology of sunfishes: patterns of variation in the feeding mechanism. Zool. J. Linn. Sot. 1986: 217--22X.

Wcrncr, E. E. 1977. Species packing and niche complementarity in three sunfishes. Amer. Nat. I I I: s53 57x.

Werner. E. E. and D. J. Hall. 1979. Foraging etliciency and habitat switching in competing sunfishes. Ecology 60: 256 264.

Winterbottom, R. 1974. A dcscriptivc synonymy of the striated muscles of the teleostci. Proc. Acad. Nat. Sci. Phil. 125: 225-317.

Wittc, F. 1984. Ecological differentiation in Lake Victoria Haplochromines: Comparison of clchlid species llocks in African Lakes. In A. A. Echelle and J. Kornfield (Eds.). Evolution of Fish Species Flocks. University of Maine and Orono Press, Orono, Maine.

Wittc, F. and C. D. N. Barel. 1976. The comparative functional morphology of the pharyngcal .jaw apparatus of piscivorous and intrapharyngcal mollusc-crushing Iftrploc~hromis spccics. Rev. Trav. Inst. Ptches marit. 40: 793-796.

Witte. F.. T. Goldschmidt, J. Wanink. M. van Oijen, K. Goudswaard. E. Witte-Maas and N. Bouton. 1992. The destruction of an endemic species flock: quantitative data on the decline of the

haplochromine cichhds of Lake Victoria. Environ. Biol. Fish. 34: I 2X. Yamaoka, K. 1978. Pharyngeal jaw structure in labrid fishes. Publ. Seto Mar. Biol. Lab. 24: 409 426. Yamaoka, K. 1991. Feeding relationships. pp. I51 172. Itr M. II. A. Keenlcysidc (Ed.), Cichlid Fishes.

Bchnviour, Ecology and Evolution. Chapman and Hall. London

Rccclvcd 4 October 199.5; acccptcd I February 1996. Corresponding Editor: C. D. Schlichting

Pharyngeal biting mechanics in centrarchid and cichlid ...frietsongalis.nl/wp-content/uploads/2009/06/Galis-Drucker-1996-J...9516, NL-2300 RA Lriden, ... Pharyngeal biting in perciform

Documents