Contributions to Zoology, 72 (I) 17-37 (2003)
SPB Academic Publishing bv, The Hague
The clavobranchialis musculature in sarcopterygian fishes, and contribution to osteichthyan feeding and respiration
Zerina Johanson
Palaeontology, Australian Museum, 6 College Street, Sydney, NSW 2010, Australia, e-mail: [email protected]
Keywords:: Clavobranchiales, Sarcopterygii, Actinopterygii, Chondrichthyes, coracobranchiales, Dipnoi,
Neoceratodus, Strepsodus
Abstract Introduction
Various fossil distinct the lungfish taxa preserve depressions on Muscles originating on pectoral girdle and in- the smooth postbranchial lamina ofthe dermal pectoral girdle. serting anteriorly on the mandible and branchial These depressions are largely unknown in other sarcopterygian arches depress or lower the mandible, hyoid arch, fishes, are but present in the rhizodont sarcopterygian Strepsodus. and more posterior gill arches. Of these muscles, Comparisons extant fishes these with actinopterygian suggest
much of the research into ver- depressions mark the point of origin for the clavobranchialis gnathostome (jawed and has musculature, extending anterodorsally into the gill chamber to tebrates) feeding respiration focused on the
insert on the ventral surface ofthe Studios ceratobranchial(s). sternohyoideus and coracomandibularis. The sterno- examining feeding and respiratory mechanisms ofbony fishes hyoideus (= rectus cervicus) attaches to the ventral (Osteichthyes) have emphasised the role of mandibular depres- of the hyoid arch and lowers the mandible within the oral portions sion in generating negative pressures cavity to draw in via the the latter run- water/air/food via suction. However, phylogenetically mandibulohyoid ligament, basal dorsal and actinopterygians, fossil lungfish and other fossil sarcoptc- ning between the ceratohyal the rear of
rygians (such as Strepsodus) lack the apomorphies that increase the mandible (Lauder, 1979, 1980, 1982, 1983a, b, suction among bony fishes. In these taxa the clavobranchialis 1985; Lauder& Shaffer, 1985; Bemis, 1987; Bemis muscles to this re- may serve augment negative pressure by & Lauder, 1986; Shaffer & Lauder, 1985; & the Reilly tracting ceratobranchialsand increasing the size of the oral/
Lauder, 1990). This was described as a oropharyngeal cavity. A comparable action is performed by the general- ised the chondrichthyan coracobranchiales muscles, particularly during mechanism for Gnathostomata, occurring
feeding, and the is function of these ventral gill arch muscles in acanthodians, actinopterygians, extant lungfishes, likely to be a synapomorphy of jawed vertebrates (Gnatho- coclacanths (the latter two groups representing the stomata). This musculature is absent from jawless vertebrates such Sarcopterygii) and aquatic salamanders as the Osteostraci. piscine (Lau- der, 1980, 1982, 1985; Bemis & Lauder, 1986;
Campbell & Barwick, 1987, 1988a, 1999; Maisey,
Contents 1989; Reilly & Lauder, 1990). Attachment surfaces
for the sternohyoideus muscle and mandibulohyoid
Introduction are also in various fossil sarco- 17 ligament preserved
Materials ‘ and methods 18 pterygian taxa including Eusthenopteron (Jarvik,
The dermal pectoral girdle of fossil sarcopterygians and 1980), Gogonasus (Long et ah, 1997: figs 39, 48B,
associated muscle attachments 19 D), Medoevia (Lebedev, 1995: fig. 18B), The Glyptolepis role of the clavobranchialis in feeding and respira- 1972: and several fossil tion (Jarvik, fig. 28), lungfishes 2'j 1977; et ah, 1993; & Bar- Discussion 30 (Miles, Wang Campbell
Conclusions Barwick & 33 wick, 1988a, 2002; Campbell, 1996;
Acknowledgements ' KSW 33 Campbell, pers. comm. 2002). However, re- References 34 cent research on the chondrichthyan musculature 18 Z. Johanson - Clavobranchialis musculature in fossil lungfishes
questioned the function ofthe mandibulohyoid liga- (Lauder, 1979, 1982, 1983a, b; Bemis & Lauder,
in this the main muscle fossil ment group. Rather, lowering 1986). Nevertheless, among sarcopterygians, the mandible was the coracomandibularis, inserting a variety of fossil lungfish possess well developed at the symphysis of the chondrichthyan lower jaw muscle attachment surfaces on the postbranchial
(Wilga et ah, 2000). laminae of the pectoral girdle. Comparable attach-
As the mandibleis loweredand the mouth opened, ment surfaces occur on the clavicles of the rhizodont negative pressure (suction) created within the oral sarcopterygian fish Strepsodus. These surfaces may
draws in aerated and food indicate the of clavobranchiales cavity water (or air) presence a com- materials. Mandibular depression was said to cre- parable to various actinopterygians (Jessen, 1972),
in ate the greatest change in the volume of the oral and the size of the depression lungfish such as cavity and provide the greatest contribution to this Griphognathus indicates that the muscles themselves
could be of substantial size. negative pressure (Lauder, 1985). Additionally, actinopterygians (e.g., Alexander, 1969, 1970; Liem, As described below, the action of ventral gill
arch have both 1970; Lauder, 1983b), lepidosirenid lungfish (Pro- depressors may improved feeding topterus and Lepidosiren (Bemis, 1987; Bemis & and respiratory efficiency by increasing the rate at
Lauder, 1986)) and aquatic salamanders (Lauder which water/food/air is brought into the oral cav-
& Shaffer, 1985) modify various aspects of their ity in chondrichthyans, sarcopterygians and actino- jaws, skull, gill arch, pectoral girdle and muscle pterygians, and particularly in phylogenetically basal
increase of the latter Evidence for the morphology to negative pressure gener- taxa two groups. pres- ated within the oral cavity. These modifications serve ence of these muscles and active branchial arch to increase the flow of water/air into the mouth, depression can also be found in certain fossil gnatho- improving the efficiency of both respiration and stomes such as the Placodermi, suggesting the ac- feeding. By comparison, fossil sarcopterygians (in- tion of this gill arch musculature in feeding and cluding non-lepidosirenid lungfish) appear to lack respiration occurs throughout jawed fishes. By com- these modifications, as do fossil and phylogenetically parison, the action of the sternohyoideus muscle in basal actinopterygians (Lauder, 1980, 1982), and lowering the mandible and opening the mouth the within the mouth via the arch and negative pressure generated hyoid mandibulohyoid liga- cavity of these taxa was considered to be relatively ment, described as a feature of the Gnathostomata low (Lauder, 1980). (e.g., Lauder, 1980, 1982), does not occur in chon-
However, little consideration has been given in drichthyans (Wilga et ah, 2000), nor in placoderms these discussions to the musculature responsible (Johanson, in press). In these taxa the coracomandi- for the more and the the depressing posterior gill arches, bularis, inserting at jaw symphysis, opens its role in expanding and increasing the volume of mouth. the oropharyngeal cavity during the initial stages of feeding and respiration. These muscles, in- cluding the clavobranchiales in osteichthyans and Materials and methods the coracobranchiales in chondrichthyans, originate on (or in association with) the pectoral girdle and Specimens illustrated were either photographed after insert on the ventral the surfaces of branchial arches being coated with an ammonium chloride sublimate
evidence indi- scanned in (ceratobranchials). Experimental (Figs 3, 4, 5) or from various publica- cates that the coracobranchialesact chondrichthyan tions (Figs 1, 2, 7). Line drawings (Fig. 6) were to the branchial basket depress during feeding made directly from photographs. In Figure 8, a 25
(Moss, 1977; Mallatt, Motta 1997: 1996; et ah, cm long individual of the lungfish Neoceratodus lig. 4), a link the indicating between activity of was cut by hand into sections and photographed these muscles and the movement of and Institutional water food on a light table. abbreviations: AMF: into the mouth. Similar of activity the clavobran- Australian Museum, Sydney; QMF: Queensland chialis has musculature not been examined in sarco- Museum, Brisbane. pterygians (extant lungfish) nor in actinopterygians Contributions to Zoology, 72 (I) - 2003 19
The dermal pectoral girdle of fossil and includes the Porolepiformes, inwhich the lamina
sarcoptery-gians and associated muscle can be relatively wide with no indication of mus-
attachments attachments cle present (Jarvik, 1972). Youngolepis
(Chang, 1991) is also assigned to the Dipnomorpha The dermal pectoral girdle of fossil sarcopterygians (Ahlberg, 1991) and its postbranchial lamina ap-
is well known be covered in dermal (e.g., Andrews & Westoll, 1970a, b; pears to ornament (also the
Jarvik, 1972, 1980; Miles, 1977; Long, 1989; Fox onychodont Strunius) rather than smooth. The Tetra- et ah, 1995; Lebedev, 1995; Ahlberg & Johanson, podomorpha include taxa more closely related to
Johanson than 1997; & Ahlberg, 1997). Among these, a the Tetrapoda the Dipnoi (lungfish) and gen-
of reduced variety fossil lungfish including Sagenodus, Eoc- erally possess postbranchial lamina, often
tenodus, Pillararhynchus, Holodipterus, Gripho- only represented by a narrow, unornamented strip
gnathus, Chirodipterus and Ctenodus (Watson & along the dorsomedial margin of the cleithrum and
Gill, 1923; Miles, 1977; Long, 1987; Pridmore et clavicle (Jarvik, 1972, 1980; Fox et ah, 1995; Le- ah, 1994; Barwick & Campbell, 1996; Campbell bedev, 1995). However, distinct depressions are & associated with the unornamented Barwick, 1999) possess a broad and well-devel- postbranchial
oped postbranchial lamina associated with the clei- lamina on the clavicle of Strepsodus. Strepsodus is
thrum memberof the of (cleith) and clavicle (clav), the main bones a Rhizodontida, a group tetrapodo- of the pectoral girdle. On this lamina (pbl), dis- morph fishes known primarily from teeth, scales,
tinct depressions are visible, variable in size, depth and the robust bones of the pectoral girdle (Andrews, and shape, but deepest laterally, where the lamina 1973, 1985; Andrews & Westoll, 1970b; Long,
joins the Johanson & external part of the cleithrum and clavicle 1989; Jeffery, 2001; Ahlbcrg, 1998, (Figs of from 1-3). In Pillararhynchus, Holodipterus and 2001). New material Strepsodus sp. Queens-
Eoctenodus, this depression is represented by a small, land, Australia (Johanson et al., 2000) is disarticu- shallow circle abutting the lateral margin of the lated but relatively well preserved, including several
lamina of the cleithrum (Fig. 1 A, C, white arrow). clavicles (Figs 4,5). These support a rounded de- There does not to be de- of moderate size, located between appear a corresponding pression midway
pression on the postbranchial lamina of the clavi- the base of the ascending process (as.pr) and the cle. In other the bone. The lungfish taxa, the depressions are deeper, anterior margin of depth of this de- Jre located on both the cleithrum and clavicle, and pression varies among the clavicles, but is deepest
can contain within and 4A, att. narrow, bony processes (Chiro- on QMF36954 QMF36760 (Fig. B, dipterus, Fig, These to where it is also surroundedby a distinct rim. —, 2A-C). depressions appear clavo), reach their maximum depth in Griphognathus (Fig. The base of the depression is positioned relatively 1 D). Among extant lungfishes, Neoceratodus also posteroventrally. In Figures 4 and 5, the clavicle is
possesses a robust shown in life such that the pectoral girdle and a relatively position, ascending process bioad postbranchial lamina but lacks distinct dc- of the clavicle (best preserved in QMF37408, Fig. pressions (Gunther, 1871); the girdle and laminae SA) overlaps the cleithrum posterodorsally (Fig. 6). aie reduced as a whole in Protopterus and Lepi- In this position, the depression on the clavicle has dosiren, assigned to the family Lepidosirenidae a dorsal and anterior orientation (as indicated by (Bemis & Lauder, 1986). Thus, no living lungfish the white arrows, Fig. 4C, D), becoming shallower
would and wider in this direction. The cleithrum appear to possess these depressions on the lacks a postbranchial lamina of the pectoral girdle. comparable depression (Johanson et ah, 2000). f omparable depressions on the dermal pectoral Given these morphologies in Strepsodus and the gudle are also unknown in various fossil it seems reasonable most other, non-lungfish, lungfish, to sug- sarcopterygian taxa, including coelacanths (Forey, gest that these depressions on the clavicle and/or
JM, 1998), cleithrum muscle attachments onychodonts (Jessen, 1966) and mem- represent (e.g., as
ers l * le Dipnomorpha and Tetrapodomorpha described for Sagenodus (Watson & Gill, 1923)). sensu Ahlberg (1991)), The Dipnomorpha includes 'The deepest part of the depression represents the uxa nioie closely related to lungfish than tetrapods, base of the attachment, with the muscle widening 20 Z Johanson - Clavobranchialis musculature in fossil lungfishes
Fig. I. A-D, dermalpectoral girdles of fossil lungfish, cleithra and clavicles. A, B, Pillararhynchus longi WAM 86.9.595, A, anterior, lateral B, views showing postbranchial lamina and lateral depressions for attachment of clavobranchialismusculature (from Barwick
& Scale = 1.0 et 1994: whitei Campbell, 1996). cm. C, Holodipterus meemannae (from Pridmore al., fig. 82b). D, Griphognathus
AND anterolateralview = for muscle 49117, (from Campbell & Barwick, 1999). Scale 1.0 cm. Arrows in A, C, D indicate depressions
attachment. Abbreviations; clav, clavicle; cleith, cleithrum; pbl, postbranchial lamina. Images in Fig. 1A, B used with permission of
Schweizerbart-Publishers, Stuttgart (PaleontographicaA). Image in Fig. 1C from Pridmore et al. (1994), ‘Morphology and phylogenetic
position of the holodipteran dipnoans ofthe Upper Devonian GogoFormation ofnorthwestern Australia’. Phil. Trans. R. Soc. London
B 344: 105-164, used with permission ofThe Royal Society London, and Dr. Peter Pridmore. Images in Fig. ID used with permission of Prof. Ken Campbell. Contributions to Zoology, 72 (1) - 2003 21
Fig. 2. A-C, dermalpectoral girdles offossil lungfish, cleithra and clavicles. A- B, Chirodipterus australis, , ANU35636, A, anterolateral, B, anterior views showing postbranchial lamina and lateral depressions (large white arrows in A-C) for attachment of clavobranchialis CUlatUre (from = ™l Campbell & Barwick, 1999). Scale 1.0 cm. C, Chirodipterus australis, ANU49200 (from Campbell & Barwick, 1999). Abbreviations in Figure 1, also oa.clav, overlap surface on the cleithrum for the clavicle. Images in Fig, 2A-C used with permission of Prof. Ken Campbell.
4nd possibly dividing into separate slips as it ex- 4, 5). Overall, the muscle attachments on the clei- tends outwards from this base. On Strepsodus, the thrum and/or clavicle of both the lungfishes and orientation of the muscle attachment is dorsal and Strepsodus would have been located medial to the anterior. In the the case of the lungfish taxa, anterior opercular and subopercular bones and muscles views of the cleithra and clavicle indicate that the themselves oriented anteriorly, dorsally, and also attachments would also to have medial into the chamber. appear a medially gill orientation because the deepest part of the attach- In extant actinopterygians, the clavobranchialis ment is laterally positioned (Figs 1 A, B, 2). How- musculature originates dorsally on the anterolateral evei, in surface of the on the smooth Griphognathus (Fig. ID) the depressions cleithrum, or unorna- aic cxtr emely deep and oriented almost directly mented portion representing the postbranchial lamina r oisoventrally, which Lauder & may indicate a slightly dif- (Jessen, 1972; Jollie, 1982; Liem, 1983; CICnt °'ienation and relative to other lungfishes. The Fig. 7, clavo), inserts on the ventral surface of presence of the depressions on both the cleithrum and ceratobranchial(s). This dorsal and lateral at-
0 av ' c c indicate the ' may attachment of different tachment on the dermal pectoral girdle seems to muscles in these lungfish taxa, though the contigu- compare closely with the location of the depres- ous nature of the depressions/attachments (e.g., Figs sions on the cleithra and clavicles of Strepsodus,
’ su ggests these muscles had essentially simi- Chirodipterus, Holodipterus, Griphognathus and ar unct '°ns, and were subsets of the same muscle Pillararhynchus, as described above. The antero- group. On Holodipterus, Pillararhynchus, dorsal orientation of these pectoral girdle depres- Eocteno-dus and also Strepsodus, it that sions and the clavobranchiales appears a single actinopterygian are
iiusclc mass was present, attached to the cleithrum also very similar, that is, towards the branchial arch and clavicle, respectively (Long, 1987; Figs I A-C, chamber. Flowever, in extant lungfishes, the clavo- 22 Z. Johanson - Clavobranchialis musculature in fossil lungfishes
preserved in section A originates from the ventro-
medial edge ofthis anteriorpart of the pectoral girdle
and runs dorsomedially to the ventral portions of
the branchial arches. Section B (Fig. 8B) is located
just posterior to section A and shows that the pec-
toral girdle is becoming thicker and is extending
posterodorsally. This section of the pectoral girdle
in represents the cleithrum, while the girdle sec-
A the tion represents clavicle. The gill arches are
visible in section B, but there are no muscles run-
ning from the pectoral girdle to the base of these
arches. The muscles in this region are strictly trans-
and oriented towards the midline. verse, Therefore,
the clavobranchiales in Neoceratodus are largely
restricted to the anterior and ventral portions of the
pectoral girdle, as suggested by Edgeworth (1926).
This differs from the fossil lungfish described above,
in which the depressions for muscle attachment were
located more dorsolaterally (comparable to the por-
tion of the pectoral girdle shown in Fig. 8B). These
differences may be related to a reduction in the size
of the gill arches in extant lungfish; although these
arches are large in Neoceratodus(e.g., Bemis, 1987),
they are reduced relative to the size of those in taxa Fig. 3. SagenodusAMF 6285, anterolateral view ofcleithrum, such as Griphognathus, particularly the posterior larger arrows indicate depression for clavobranchiales. Scale = arches (e.g, Campbell & Barwick, 1987: 1.0 cm. compare fig. 22, 1999 and Bemis, 1987: fig. 3B).
Another possibility, given the position and orien- branchiales musculature was said to originate on tation of the muscle attachments, is that the mus- the ‘inner and ventral edges’ of the pectoral girdle cles attaching to the pectoral girdle in Strepsodus (Edgeworth, 1926; Wiley, 1979; Jollie, 1982). Five and the fossil lungfish described above were asso- clavobranchiales are present, with the anterior four ciated with the the operculum. However, opercu-
from a and the fifth from slips arising single origin are lar and subopercular bones moved by muscles a second origin justposterior to the first (Edgeworth, originating dorsally on the endocranium (dilator
1926). Transverse sections cut through an individual adductor levator operculi, operculi, operculi, e.g., of Neoceratodus 25 (approximately cm long) con- 5 Jarvik, 1980: 95, 96). They are not likely to be firm the relatively ventral and internal of origin muscles associated with the pectoral fin, as these these muscles In section A is the (Fig. 8). Figure 8, generally originate on the scapulocoracoid rather
more anterior. The dorsal of the branchial the portions than on the anterior parts of pectoral girdle, arches are fused to the braincase, while the ventral and are oriented posteriorly to reach the fin, rather
down into the 2-5 hang gill chamber. Arches are than anterodorsally (e.g., Andrews& Westoll, 1970a: visible in this section, with the fifth being the smal- figs 31,32; Janvier, 1980; Janvier & Marcoux, 1976; lest, and hidden from view. nearly The pectoral girdle Fox et ah, 1995). Other muscles associated with can be seen on cither side of the the section, extending gill arches (e.g., transversi ventrales) run be-
and vcntrally medially towards the observer. Al- tween the different elements of the arches them-
the curvature of though cut, the preserved girdle selves, and function to constrict or contract these halves (which meet in the eventually midline) sug- elements rather than depress them (Wiley, 1979; the anterior of the gests parts girdle are visible in Jollie, 1982). The sternohyoideus and coracoman-
section A. The clavobranchiales muscle (clav'o) dibularis muscles run from the jaw symphysis or Contributions to Zoology, 72 (I) - 2003 23
Fig. 4. Clavicles of from Carboniferous Ducabrook central Australia. External Strepsodus sp. the Lower Formation, Queensland, -°f left clavicles, arrow indicates orientation of clavobranchialis muscle. A, QMF36760; B, QMF36954; C, QMF34606; D,
46u». Scale bar = 1 Abbreviations: ofclavicle onto cm. ant, anterior; as.pr, ascending process overlapping cleithrum; att.clav, pression marking the attachment surface for the clavobranchialis musculature. White arrows in Figure 4C, D indicate suggested orientation ofthe clavobranchiales, anterodorsally into the gill chamber.
arch to the ventral it is difficult pectoral girdle, but arc Thus, to come to any other conclu- line muscles & 'I" (e.g., Bemis Lauder, 1986), and ,sion as to the identity of the muscles attaching to o not originate lateral dorsally or laterally on the pecto- the depressions on the pectoral girdle of ral girdle. various fossil lungfishes and the rhizodont Strepso- - 24 Z. Johanson Clavobmnchialis musculature in fossil lungfishes
Fig. 5. Clavicles of from the Lower Carboniferous Ducabrook Formation, central Australia. external Strepsodus sp. Queensland, A, C,
view of right clavicle, B, external view ofleft clavicle. A, QMF37408; B, QMF37536,arrow indicates orientation of clavobranchialis
muscle; C, QMF37566. = = Scale bar 1 cm, except for B, C, where scale bar 0,5 cm. Abbreviations as in Figure 4.
dus. By comparison to extant actinopterygians (e.g., culature, as described above for the extant lungfishes
these the of four the clavicle the Fig. 7), depressions indicate point at- (first slips anteriorly on and
tachment of the clavobranchiales (origin) muscles. fifth posteriorly on the cleithrum). Holodipterus,
In Strepsodus, the of a muscle is and Eoctenodus have resembl- presence single Pillararhynchus may in the indicated, as actinopterygians. In fossil lung- ed the condition in Strepsodus and the extant actino-
fish such as and Griphognathus Chirodipterus, de- pterygians illustrated in Fig. 7, with a single muscle
pressions on both the cleithrum and clavicle indicate origin on the cleithrum (Fig. 1). This muscle could
two for the clavobranchiales separate origins mus- have comprised several slips as in Neoceratodus, Contributions to Zoology, 72 (I) - 2003 25
cle attachments comparable to those described above
for the fossil forms are lacking (Gunther, 1871;
Jarvik, 1980; pers. obs.). Nevertheless, anterior and
posterior clavobranchiales are present, attaching to
all five gill arches, although the latter arches are
relatively reduced in size (Edgeworth, 1926; Wiley,
1979; Bemis, 1987).
The above discussion focused on previously un-
described (fossil lungftsh) and new (Strepsodus)
features on the dermal pectoral girdle indicating
the attachment of origin or muscles; these were
the identified as clavobranchiales by comparison
to extant actinopterygians. Interpreting the phylo-
genetic significance of this character distribution
is more problematic, since evidence for the attach-
ment of this musculature in the form of pits or
depressions is absent from the postbranchial lamina
F‘S- 6. A, of B, QMF36954, line drawing. B, reconstruction of most fossil sarcopterygian taxa (Jessen, 1966; of clavicle and Strepsodus sp, (based onQMF36954 QMF37408) Jarvik, 1972, 1980; Forey, 1981, Fox et ah, 1995; and its relationship to the rhizodont cleithrum(e.g., 1989)). Long, Lebedev, 1995). Most of these taxa have smooth,
if reduced, postbranchial laminae, but in other taxa, but a separation into anterior and posterior muscle the laminae are covered in ornamentation, precluding masses of muscle attachment. appears not to have occurred. Branchial any possibility The latter ‘irches are not well preserved in fossil lungfishes, include the onychodont Strunius (Jessen, 1966) and but in Griphognathus, all arches are large and well the dipnomorph Youngolepis (Chang, 1991), as well formed, from anterior to posterior, with no sub- as phylogenetically basal actinopterygians such stantial decrease obs. in size posteriorly (Miles, 1977; as Mimia (pers. AMF119672) and Dialipina Campbell & Barwick, 1987: figs 21,22). Branchial (Schultze & Cuumba, 2001). ‘" ches (ceratobranchials) are also known for Chiro- The available evidence suggests that a dorsola-
dipterus and Holodipterus (Miles, 1977), where teral origin of the clavobranchialis musculature on a gam, they appear relatively well developed, al- the pectoral girdle occurs in tetrapodomorph and
110ugh a complete set of arches, comparable to Gri- dipnomorph sarcopterygians ((Strepsodus and fos- phognathus, is not preserved. As described above, sil lungfish, respectively) and actinopterygians (Jes- 1 e muscle for the of the In depressions attachment sen, 1972). chondrichthyans, by comparison, the 0 av °branchiales in Chirodipterus are large, occur gill arch depressing coracobranchiales muscles ori- mi both bones ot the and the pectoral girdle, may have ginate in the midline, ventrally on coracoarcualis icld bony splints for additional muscle attachment, muscle (which extends anteriorly from the ventral be muscle the coracoid attachment on the pectoral girdle of part of (e.g., Motta & Wilga, 1999)). Holodipterus is much smaller by comparison, and Although this evidence is incomplete, it distinguishes one could predict that a set of gill arches similar to osteichthyans (sarcopterygians and actinopterygians)
Griphognathus was present in Chirodipterus, with from chondrichthyans. Actinopterygian taxa such some reduction of these in the have Holodipterus. Flowevcr, extant genusAcipenser a wide postbranchial a till set of arches would be this but the clavobranchiales needed to test lamina, bypass this sur- suggestion. It is also somewhat difficult to make face altogether to insert on the scapulocoracoid generalisations as to arch size and the 1972: which also gill corre- (Jessen, pi. 5), may be the case sponding size ot the clavobranchiales; Neocerato-in for the sarcopterygians and actinopterygians lack- dus, the postbranchial lamina associated with the ing evidence for the attachment of the clavobran- e >t rum and clavicle is well-developed, but mus- chiales on the dermal pectoral girdle. This could - 26 Z. Johanson Clavobranchialis musculature in fossil lungfishes
7. Fig. Actinopterygian gill arch musculature, lateral view ofcleithrum and muscles, including clavobranchiales (clavo) passing into
gill chamber (adapted from Jessen, calva. 1972: figs 7.3, 8.3, 9.3), A, Elops saurus. B, Amia C, Lepisosleus osseus. Images used with permission of Taylor & Francis (Fossils and Strata).
still be considered a dorsal relative the The fossil fish Placodermi is origin to jawed group gen- condition in chondrichthyans, where the coraco- erally resolved phylogenetically to the base of the
branchiales are restricted to a ventral and midline cladeGnathostomata (Janvier, 1996,2001). Among
origin. certain placoderms, taxa possess depressions on the Contributions to Zoology, 72 (1) - 2003 27
branchiales, attaching to a ventral midline muscle
like the chondrichthyan coracoarcuales.
Given these it is observations, interesting to con-
sider the pectoral girdle of the sarcopterygian (or
basal osteichthyan) Psarolepis (Zhu et al., 1999;
Zhu & Schultze, 2001). This pectoral girdle shows
bone in an extra comparable relative position to
the spinal plate of the placoderm trunkshield. The
postbranchial lamina of the Psarolepis pectoral
girdle is covered in ornament (Zhu et al., 1999:
fig. 2), and no muscle attachment surfaces are vis-
ible on this lamina. Furthermore, the scapulocoracoid
is flat, and closely attached to the internal surface
ofthe similarto the condition girdle, very in a variety
of placoderms (e.g., Stensio, 1959; 0rvig, 1975;
Young, 1980). The scapulocoracoid in Psarolepis
have been hidden behind may largely the pectoral
of girdle, preventing attachment a clavobranchiales
musculature, and the ventral gill arch depressors
may have also been more comparable to the chon-
drichthyan coracobranchiales. Thus, although there
is a broad, potentially phylogenetically meaning-
ful distribution regarding the origin of the ventral
the bothof origins characterising these groups would anterolateral in the fossil margins of the trunkshield compara- appearto be present group Placodermi. 1 e >n position to those on the postbranchial lami- nac ofStrepsodus and the fossil lungfish described 3 ove anb the to position of the clavobranchiales The role of the clavobranchialis in feeding and muscle in extant actinopterygians (Heintz, 1932, respiration
1975; Dennis & Miles, ' ' a , 1979b; in Johanson, press). These depres- Initial stages in feeding and respiration involve sions appear to characterise the derived aerated water/air/food into the only placo- bringing mouth or erm 8r Arthrodira it oup (Goujet, 2001), and is oral cavity. The generalised osteichthyan (Sarco- suggested this group may have possessed a clavo- pterygii + Actinopterygii) mechanism by which this • anchiales musculature depressing the branchial occurs includes a lowering of the mandible via the niches. In other placoderm taxa, these depressions mandibulohyoid ligament through the posteroventral are absent ; as well, the trunkshield covered the depression of the hyoid (by the sternohyoideus scapulocoracoid and that laterally anteriorly, so a muscle; Lauder, 1980, 1982, 1983a, 1985; Bemis orsolateral muscle attachment the on scapuloco- & Lauder, 1986). Depression of the hyoid results acoid (comparable to the for in movement suggested position a posterodorsal of the dorsal part of ic sarcopterygians and actinopterygians described the ceratohyal, transmitted through the mandibulo- ° woubl have been In the the impossible. these placo- hyoid ligament to rear of jaw, which is rotated c | riT1 taxa Ihc ’ ventral arch gill depressors have around the These actions k may quadrate-articular joint. cn more comparable to the chondrichthyan coraco- open the mouth and expand the oral cavity, creat- 28 Z. Johanson - Clavohranchialis musculature in fossil lungfishes
a within which in smaller ing negative pressure (suction) negative pressures in the oral cavity, and
turn results in the flow of water (or air) and food presumably this would also have been true of fossil
into the mouth. Chondrichthyans possess a mandi- sarcopterygians, including non-lepidosirenid lung-
bulohyoid ligament, but electromyographic and fish.
manipulation experiments have indicated that the Lauder (1980) also discountedthe suggestion that
action of the ligament is decoupled from that of posteroventral retraction of the branchial arches
the the would result in coracohyoidcus (= sternohyoideus; Wilga larger negative pressures in the oral
et ah, 2000). Instead, the chondrichthyan mandi- cavity. However, experimental evidence indicates
ble is the action depressed by of the coracoman- that the coracobranchiales muscles of sharks are
dibularis muscle. All in and of these muscles involved active during the feeding strike to a lesser extent
the depressing the mandible and/or hyoid arch are during respiration (Motta et ah, 1997; Mallatt, 1996)
of the muscles part hypobranchial musculature, originating and that modifications in these can im-
the the the on pectoral girdle near midline, on hypax- prove the efficiency of suction feeding in certain
ial musculature or on the pericardium surrounding sharks (Moss, 1977; Motta & Wilga, 1999). This
the heart & (Daniel, 1934; Jessen, 1972; Bcmis evidence links muscle activity and depression or
Lauder, 1986; Motta & Wilga, 1995; Wilga & Motta, retraction of the branchial arches with feeding and
and indicates that ofthe 1998). respiration movement pos-
The focus of this discussion is on the mechanics terior branchial arches increases negative pressures
of actively drawing air/water/food into the oral cavity within the oral cavity. This is perhaps not surpris-
the during feeding and respiration. Generally, studies ing, since retracting or depressing posterior gill
on this aspect of respiration and feeding have con- arches must increase the overall size of the oral
centrated on the actions of the main hypobranchial and pharyngeal cavities.
muscles opening the mouthand expanding the oral For example, halecostome actinopterygians, in-
but have said little about contribu- Amia and teleost cavity, possible cluding fishes, are able to pro-
tions of the clavobranchiales (bony fishes) or coraco- trude the maxilla (Lauder, 1980, 1982, 1985; Liem,
branchiales (in chondrichthyans) attaching to and 1980; Lauder & Liem, 1983), creating a rounder
and depressing the ventral gill arches. One reason for more elongate, tube-like mouth opening, and this may be that the clavobranchiales (= pharygno- increasing the rate of flow ofwater into the mouth.
clavicularis and interims Extant such externus (Jollie, 1982)) lungfishes as Lepidosiren possess com-
function in food in the manipulation Euteleostei, plex overlapping and interlocking upper and lower
a large, phylogenetically derived actinopterygian lips (absent in Neoceratodus; Bemis, 1987: 263)
characterised sizeable denticulated which narrow the of the mouth achieve group by plates gape to the
on the last branchial arch (to which the clavobran- same effect and prevent water from escaping later-
chiales attach (Liem, 1970; Lauder & Liem, 1981; ally (Lauder, 1985: fig. 12.6). A variety of fossil
Lauder, 1983)). Additionally, several of these studies lungfish have large lips and could have narrowed
have investigated morphological specialisations \the mouth opening (Campbell & Barwick, 1987, designed to increase the efficiency of mouth ex- 1988b; Barwick & Campbell, 1996), but there is
and suction of air aerated in pansion food, or water no evidence that these interlocked as Lepidosiren.
into the oral These charac- in halecostome cavity. specialisations Additionally actinopterygians, a terise derived actinopterygians (Alexander, 1969, second jaw-lowering mechanism acts through the
1970; Liem, 1970, 1980; Lauder, 1980, 1982; Lauder opercular series of bones and the retraction of the
& Liem, 1983), lepidosirenid lungfish ((Protopterus levator operculi (Alexander, 1969; Liem, 1970,
+ Lepidosiren ; Bcmis, 1987; Bemis & Lauder, 1986; 1980; Lauder, 1980). Here, the interopercular bone
Campbell & Barwick, 1988b), tetrapods (Schaffer (absent in non-halecostome actinopterygians) trans-
& Lauder, 1985), and certain sharks (Moss, 1977; mits the movement of the levator operculi through Motta & Lauder Wilga, 1999). (1980: 315) sug- to the mandible. This acts alongside the jaw mecha- gested that plesiomorphic actinopterygians lacking nism acting through the mandibulohyoid ligament these features specialised would have generated and increases the speed at which the mandible is Contributions to Zoology, 72 (1) - 2003 29
lowered and the mouth opened. Lepidosirenid lung- effects of the posterior retraction by the hypaxial
the of fishes and aquatic amphibians (characterised as musculature would be offset by size the suction and its solid attachment the skull. feeders) possess a depressor mandibulae pectoral girdle to muscle involved in depressing the jaw, increasing In the lepidosirenid lungfish (Lepidosiren + Proto- the speed at which the mouth is opened (Bemis, pterus), the pectoral girdle is substantially reduced
in 1987; Bemis & Lauder, 1986; Gillis & Lauder, 1994; size, and potentially more readily retracted (Be-
Lauder & Shaffer, 1985; Reilly, 1995, 1996; Reilly nds, 1987). Additionally, the anocleithrum is absent
& Lauder, 1990). (Owen, 1841; Bishop & Foxon, 1968; McMahon,
Lauder (1980) noted that the features charac- 1969; Bemis, 1987), and the pectoral girdle, freed terising the halecostome actinopterygians were ab- from its connection to the skull, is stabilised during
in would retraction muscular attachment the sent more primitive actinopterygians, and by to cranial also be lacking in fossil sarcopterygians. For ex- rib (Bishop & Foxon, 1968; Bemis, 1987). Cranial ample, the maxilla is firmly fixed to the skull in ribs, articulating to the occipital region of the skull,
and sarcopterygian fishes (Jarvik, 1980; Long, 1989; also occur in Neoceratodus (Gunther, 1871)
Fox & in et ah, 1995; Long et ah, 1997; Ahlberg have been recognised a variety of fossil lungfish
Johanson, 1997) and absent in lungfishes and coela- (Schultze, 1975; Long, 1993; Ahlberg et al., 2001). canths (Miles, 1977; Janvier, 1996). There is no However, only in lepidosirenids are the cranial ribs
than the trunk and evidence of an interopercular bone in these taxa. substantially larger pleural ribs, Neoceratodus lacks the lepidosirenid specialisations they articulate to the rear of the chondrocranium described above, and it is likely that most fossil via a moveable synovial articulation (Bemis, 1987). lungfish lack them as well (Neoceratodus is more The cranial rib is more similar in size to other ribs
related + than in Neoceratodus and the fossil and the closely to Lepidosiren Protopterus taxa, syno- to most fossil lungfish (e.g., Schultze & Marshall, vial articulation is absent (Gunther, 1871; Miles,
1993)). In terms of the ram-suction feeding index, 1977; Bemis, 1987). As well, Neoceratodusretains plesiomorphic actinopterygians and fossil sarco- an anocleithrum, joining the pectoral girdle to the
in pterygians (including non-lcpidosirenid lungfishes) skull (Gunther, 1871), as the fossil lungfish taxa. would be to end of the The would therefore be a placed nearer ram feeding pectoral girdle very spectrum, reflecting the absence of morphological mobile unit in lepidosirenids, particularly compared specialisations of the skull, jaws or musculature to Neoceratodus and other lungfish. When retracted and their reduced the this ability to generate negative pres- by hypaxial musculature, mobility results sures in the oral cavity. in increased retraction (and speed of retraction) of Because jaw-lowering muscles such as the sterno- the mandible and hyoid arch. This would lead to a
hyoideus and coracohyoideus originate in associa- more rapid oral expansion and a more rapid increase
tion with the in the in the oral pectoral girdle, an additional feature negative pressure cavity, whether
these influencing the speed of mouth opening is the de- lungfishes are feeding (Lauder, 1985) or re-
gree to which the pectoral girdle is retracted by the spiring by gulping air (Bishop & Foxon, 1968; hypaxial musculature (Lauder, 1985; Campbell & Campbell & Barwick, 1988b). This mobile pecto-
ral be Barwick, 1988b). In plesiomorphic actinopterygians girdle would appear to another lepidosirenid and fossil sarcopterygians, the pectoral girdle is speciality that is not as well developed or even absent large, and attached to the dorsal roof of the skull (e.g., interlocking lips, depressor mandibulae) from
via a succession ofbones (lessen, 1972; Jarvik, 1980; other lungfish taxa, including Neoceratodus.
Lauder, 1980; Lauder& there are a of Liem, 1983; Gardiner, 1984; Thus, variety morphological spe- Johanson & Ahlberg, 1997; Fig. 7). This is also cialisations shown by fishes to increase the rate at
the in case fossil lungfishes, where the girdle is which water/food/air enters their mouth during the stout (e.g., Figs 1-3) and joined to the dorsal skull retraction of the hyoid arch and mandible. Fossil toot via a well-developed anocleithrum (Schultze, sarcopterygians (including fossil lungfish and to 1977 1987; > Jarvik, 1980; Barwick & Campbell, some degree, Neoceratodus) as well as plesiomor- 1996; Schultze & Chorn, 1998). In these taxa, the phic and fossil actinopterygians would generally - 30 Z Johanson Clavobranchialis musculature in fossil lungflshes
lack these basibranchial of of appear to specialisations. Nevertheless, a species Holodipterus (Pridmore
(he ability to enhance suction forces within the oral et al., 1994), but it lacks denticles entirely (Campbell
cavities beyond those generated by depression of & Barwick, 1999: fig. 13C), and the muscle attach-
in the lower jaw these taxa should not be discounted; ment for the clavobranchiales on the pectoral gir-
this the in regard, clavobranchialis has received little dle was relatively small. By comparison, denticulated
attention. For example, Lauder (1980: 315) did not plates are absent in other taxa, including Chirodip-
believe that retraction of the branchial arches would terus, in which the clavobranchiales muscle attach-
result in in the oral a large negative pressure cav- ments were quite large (Miles, 1977; Fig. 2). Among
ity. However, he also noted that all branchial mus- non-dipnoan sarcopterygians, denticulated plates are
cles were active during the first two feeding phases present on the branchial arches of Eusthenopteron
of initial strike and oral manipulation, where the (Jarvik, 1980), but are not clearly developed on other
initial strike involved the food item into bringing tetrapodomorphs (sensu Ahlberg (1991)) such as
the oral 1983a: This is et 1997; Medoevia cavity (Lauder, 28). compa- Gogonasus (Long ah, fig. 46) or
rable to data from extant sharks indicating that the (Lebedev, 1995).
coracobranchialis muscles were active and mov-
ing the ventral gill arches during feeding (Motta et Discussion ah, 1997: fig. 4) and respiration (Mallatt, 1996).
By retracting the ventral gill arch elements, the
The mandible of fishes is lowered the clavobranchial and coracobranchial muscles expand jawed (and either action of the the branchial basket and increase the size of the mouth opened) by the sterno-
muscles (via the mandibulohyoid oropharyngeal cavity. The action of these muscles hyoideus ligament and in hyoid arch) osteichthyans or by coracomandi- supplements oral expansion caused by mandibular
and bularis muscles in chondrichthyans. taxa as- depression most importantly, would appear to Many
the to the fossil fish Placodermi occur at same time as mandibular depression, signed jawed group lack attachment surfaces for the the main activity bringing water/food/air into the mandibulohyoid
on the lower or have these in a func- oral cavity. The action of the clavobranchial and ligament jaw,
the coracobranchial musculature does not rival the ster- tionally inappropriate position. Thus, placoderm
have also been lowered coracomandi- nohyoideus/coracohyoideus in expanding the oral jaw may by a bularis in but have served to muscle inserting at the lower jaw symphysis cavity osteichthyans, may in With supplement this expansion in taxa lacking morpho- (Wilga et ah, 2000; Johanson, press). regards
to mechanisms related to feeding and respiration, a logical features improving the efficiency of water/ feature How into more characteristic be the air/food the mouth. The distinct muscu- gnathostome may
and action of the coracobranchiales and lar attachments on the dermal pectoral girdle de- presence clavobranchiales muscles in depressing the branchial scribed above in Strepsodus and in lungfishes such arches, the buccopharyngeal and as Ghphognathus and Chirodipterus indicate that expanding cavity this the suction within. The coracobranchiales was a noticeably large muscle(s), occupying increasing
the postbranchial laminaeofboth the cleithrum and and clavobranchiales muscles have a different inner-
vation (Jollie, 1982) but both to the branchial clavicle in the latter two taxa. It does not seem likely belong
musculature, while the and coraco- that these clavobranchiales functioned generally in sternohyoideus
in mandibularis are hypobranchial muscles food processing lungfishes in the manner de- (Edge- Motta & scribed for euteleostean actinopterygians, where the worth, 1935; Miyake et ah, 1992; Wilga, both last branchial arch carries opposing tooth plates 1995, 1999; Mallatt, 1996). Developmentally, muscle derive from the dorsally and ventrally (e.g., Liem, 1970; Lauder, groups paraxial mesoderm,
the branchial from the more anterior 1983a). Denticulated plates do occur on the branchial unsegmented cranial mesoderm and the arches of Griphognathus and (Miles, 1977; Camp- paraxial hypobranchial
bell & but these from the more posterior segmented somitic parax- Barwick, 1987, 1999), appear to ial mesoderm & be largely restricted to the midline basibranchials. (Schilling Kimmel, 1994, 1997; Hacker A was to & Guthrie, 1998). plate also said be present on a basihyal or - Contributions to Zoology, 72 (I) 2003 31
The presence of gill arch depressing muscles such pear to have been more active swimmers than some
as the clavobranchiales and coracobranchiales may jawed fishes, for example, various placoderms (such be a synapomorphy for the jawed fishes as they as the antiarch Remigolepis ) with relatively short
would be absent in fishes. Bran- trunks and appear to jawless caudal fins, small pectoral fins and no
chial openings (and gill arches) are present in the dorsal fins (Janvier, 1996: fig. 4.53; Johanson, 1997).
fossil fish Ventral arch muscles jawless groups Anaspida, Thelodonti, gill comparable to the osteich-
Osteostraci and Galeaspida, and in the latter two thyan clavobranchialis and/or the chondrichthyan
taxa, the position of the gill arches is clearly vis- coracobranchialis would have been useful in feed-
ible the and for on the cartilaginous walls throughout large ing, perhaps more importantly a putatively
oropharyngeal cavity (Janvier, 1996). In the Osteo- active osteostracan, respiration. Alternative expla-
straci, the pectoral girdle is anteriorly positioned, nations for the absence of these muscles are de-
and is part of the headshield, while the scapulo- scribed below, based on considerations of gill arch
coracoid the fin is located lateral and supporting on a homology gene expression in these arches in
extension of the wall of the the and how this have influenced the rear oropharyngeal cavity lamprey, may
(Janvier, 2001). Any muscles depressing the osteo- development of the branchial musculature in jawless
stracan ventral gill arches in the manner of the fishes, including fossil forms such as the osteo-
clavobranchiales/coracobranchiales would have to stracans. As well, the possible influence of the
attach to these posterior walls of the orobranchial anterior position of the pectoral fin in osteostracans
cavity. These walls separated the oropharyngeal on the branchial and hypobranchial musculature is
cavity from the trunk, precluding an attachment to discussed, based on recent evidence on muscle de- the hypaxial musculature; as well, the heart was velopment from the paraxial mesoderm in extant
enclosed within this posterior wall (Janvier, 2001), animals such as chondrichthyans, zebrafish (Neyt
How- impeding any attachment to the pericardium. et ah, 2000) and amniote tetrapods (Dietrich et al,
ever, sections through the osteostracan headshield 1999; Hacker & Guthrie). indicate that the gill arches extended across the For example, the absence of ventral gill arch de-
posterior walls of the relate of the arches oropharyngeal cavity, filling pressors may to the homology gill the entire cavity dorsally and ventrally, anteriorly betweenjawless andjawed fishes. In extantjawless and posteriorly, leaving no apparent room for muscle fishes such as the lamprey, the arch itself is lateral
attachment (Stensio, 1927; fig. 40). Intrabranchial to the gills, while in jawed fishes, the arch and its muscles contracting the gill arches have been re- related blood and nerve supply are medial to the
constructed in osteostracans (Janvier, 1985: figs 14B, gill (Schaeffer & Thomson, 1980; Janvier, 1996;
19B), and increases in the size of the oropharyn- Mallatt, 1996). In osteostracans, the gill arch is lateral geal cavity may have depended on passive recoil relative to the gills, as in lampreys (Janvier, 1985). ol these muscles during expiration, as in extant The lateral and medial positions of these branchial
arches lampreys (Randall, 1972). arches suggests these are not homologous The absence of muscles depressing the gill arches and that the ventral gill arch musculature devel-
(and lower in jaws) has been linked to functional and oped jawed fishes in conjunction with the me-
ecological scenarios involving active inspiration dial arches.
(ventilation) and increased predation or foraging However, recent considerations of shared simi-
associated with the evolution of larities in the of jaws (Mallatt, 1996). development lamprey and gna-
However, these be thostome branchial arches indicate explanations may inadequate, that these are for it is somewhat that In both example, surprising osteo- homologous (Kimmel et ah, 2001). groups, stracans lack muscles actively expanding the oral/ neural crest (ultimately forming the branchial arch)
had fins extends form oropharyngeal cavities; they pectoral sup- ventrally to a ‘shell’ around a core of
ported by moveable and associated rays an muscu- unsegmented cephalic mesoderm(forming the bran-
lature, as well as an elongate trunk, dorsal fim(s), chial musculature). In lampreys, the neural crest and a fin 1996: well-developed caudal (Janvier, fig. ‘shell’ is restricted to the lateral side of the meso- 4-14). In certain respects, osteostracans would ap- dermal core, forming the branchial arch in this 32 Z. Johanson - Clavobranchialis musculature in fossil lungftshes
In then position. gnathostome evolution, this neural crest lack a comparably restricted pattern, a ventral
continues to migrate medially around the mesoder- branchial musculature would also be lacking, in-
mal core, and there forms the medial arch charac- cluding gill arch depressors comparable to the clavo-
teristic of the The basic of the branchiales/coracobranchiales. the group. patterning Indeed, more
cephalic paraxial mesoderm and neural crest mi- posterior lamprey branchial arches have a muscu-
gration is similar in lampreys and jawed fishes, lature including external branchial constrictors, inter-
supporting branchial arch homology (Kimmel et nal dorsal and ventral diagonal constrictors and al., 2001). These observations also suggest that the median muscle bands (adductors) (Roberts, 1950;
branchial musculature both muscles is homologous in groups. Mallatt, 1996). A comparable set of is re-
One important observation involves the devel- constructed for the Osteostraci (Janvier, 1985: fig.
of dorsal-ventral in and ventral arch musculature is absent. opment a gradient gene expres- 19), a gill
sion within the gill arches of jawed fishes (Miller Another explanation for the absence of both the
et al., 2000; Kimmel et ah, 2001; Neidert et ah, clavobranchiales/coracobranchiales and the hypo-
in the 2001). This includes gradients neural crest branchial musculature in jawless fishes is based on
derived tissues for the of the anterior of the fin necessary development unusually position pectoral
in and the ventral cartilages of the branchial arches, as osteostracans, the influence of the fin on well as the joints between the dorsal and ventral muscle differentiation. As noted above, both bran- cartilages (Kimmel et ah, 2001). The dorsal-ven- chial and hypobranchial muscles derive from the tral also influences the of the do the muscles of the gradient development paraxial mesoderm, as pec-
from toral & & Le branchial muscles the cephalic paraxial me- fin (Romer Parsons, 1986; Ordahl soderm. The zebrafish sucker/endothelin 1 gene is Douarin, 1992; Schilling & Kimmel, 1997; reviewed restricted to the ventral parts of the branchial arch in Hacker & Guthrie, 1998; Dietrich et ah, 1999). and mutations in this reduce the This mesoderm in genesubstantially occurs a rostral-caudal series, ventral branchial muscles (Miller et ah, 2000: fig. with the unsegmented caudal cranial paraxial me-
7E-H; Kimmel et ah, 2001), including muscles of soderm (associated with the cranial neural crest) the more posterior arches such as the transversus giving rise to the branchial musculature (Schilling ventralis. This particular mutation also affects the & Kimmel, 1994; Hacker & Guthrie, 1998) and ventral branchial arches themselves, though the the more rostral segmented somitic paraxial meso- hypobranchial muscles (e.g., the sternohyoideus) derm contributing to the hypobranchial and pecto- do not seem to be affected (Miller et ah, 2000: fig. ral fin muscles. In chondrichthyans, cells forming
7E, F). the hypobranchial muscles originate from the an-
A dorsal-ventral in terior somites of the gradient gene expression ap- segmented paraxial mesoderm, pears to be absent in lampreys (Kimmel et ah, 2001; whereas the fin muscles originate separately, from
Neidert et ah, 2001), correlated with the absence more posterior somites (e.g., Romer & Parsons, of separate dorsal and ventral arch components, and , 1986: fig. 196). By comparison, both hypobran- joints between thesearches (Morrison et ah, 2000). ' chial and pectoral fin muscles originate from the
These latter observations focused on overlapping same anterior somites in the zebrafish (Actino- distributions of Dlx in et but this is related to separate genes gnathostomes; pterygii; Neyt ah, 2000),
in these of the muscle lampreys genes are expressed throughout the ability precursor cells to become the relevant neural crest prior to migration, with separated from the somite and mobile (also in the
no This lack of Amniota One overlap. dorsoventral gradient or (chicks; e.g. Dietrich et ah, 1999)). patterning could have also influenced branchial arch element of the genetic pathway involved in this
muscle development in the lamprey. As notedabove, mobility is Lbx-1; however, this is not expressed sucker/et-l zebrafish mutants had very poorly de- in sharks, which instead show a direct epithelial veloped ventral branchial muscles, although the contribution to fin and hypobranchial muscles (Hil- dorsal muscles were little affected. Expression of debrand, 1974; Neyt et ah, 2000; Haines & Currie, sucker/endothelin-1 was limited the ventral If and mobile muscle to parts 2001). separated precursor of the arch et If (Miller ah, 2000). lampreys do cells are a synapomorphy of the osteichthyans, then Contributions to Zoology, 72 (1) - 2003 33
the chondrichthyan condition may also characterise providing signals for the development of fin mus-
culature. jawless fishes such as osteostracans (Neyt et ah,
2000). In other words, anterior somites would con- tribute to the hypobranchial musculature, and the more posterior somites to the pectoral fin muscles Conclusions in these with latter groups, no migration of muscle precursor cells or origin from the same somite (as Research into feeding and respiration has focused characterises osteichthyans such as the zebrafish). on the muscles depressing the mandible and hyoid
More importantly, the position of the pectoral arch. Expansion of the oropharyngeal cavity via fin bud influences of the fin the of the results in development pectoral lowering mandible a negative
such that the somites medial the musculature, to pressure which in turn draws aerated water and food fin will contribute to its musculature (Christ & into the mouth cavity. Certain actinopterygians and
Ordahl, 1995; reviewed in Dietrich et ah, In modifications this 1998). sarcopterygians possess of sys- osteostracans the pectoral fin is anteriorly located, tem to improve the expansion of the cavity, the rate more or less opposite to the otic/occipital region of expansion, rate of water/air flow into the cavity of the braincase, as indicated by the relative posi- and overall suction. However, other taxa lack these tions of the otic capsules and the scapulocoracoid modifications, and absent from these discussions
1996: The anterior (Janvier, figs 4.15, 4.16). posi- has been a consideration of the role of the clavo- tion of the fin have osteostracan pectoral may re- branchialis/coracobranchialis in improving suction
sulted in the contribution of cells from the most in these taxa. Attachments for the clavobranchialis anterior somites of the segmented paraxial meso- musculature have been described in fossil sarco-
derm fin the exclusion of to muscle, to a hypo- pterygians and compared to this muscle in extant branchial musculature. Thus, a hypobranchial mus- actinopterygians. This distribution, and the size of culature attaching to the lower jaw evolved after this musculature, particularly in fossil lungfishes, the pectoral fin from became dissociated the head- suggests the clavobranchialis had a role in increas- shield and more posteriorly positioned, in jawed ing expansion of the oral cavity and osteichthyan
vertebrates. Theanterior position of the osteostracan feeding and respiration. Direct experimental evi- pectoral fin also have been for the may responsible dence from chondrichthyans (Motta et ah, 1997) absence of the coracobranchiales/clavobranchiales indicates that the coracobranchiales muscles are branchial muscles responsible for depressing the involved in expanding the branchial arches during gill arches. Hacker & Guthrie (1998) showed that feeding. Although the coracobranchiales and clavo-
pectoral branchial fin, hypobranchial and muscles branchiales have a different innervation, they both
differ in their whether markedly gene expression, suggest- derive from paraxial mesoderm, cranial and
ing different How- pathways of muscle development. unsegmented (clavobranchiales) or postcranial and
ever, they demonstrated that cells from the seg- segmented into somites (coracobranchiales). Thus, mented paraxial mesoderm grafted into the more the action of these muscles in feeding and respira-
anterior of the cranial mesoderm region paraxial tion may be a gnathostome synapomorphy, being
were able to migrate to the branchial arches and absent from jawless fishes such as osteostracans.
were incorporated into the muscles associated with these arches. Hacker & Guthrie (1998) suggested
these muscle cells were originally ‘naive’, differ- Acknowledgements
entiating and attaining their identity (and particu- lar gene the rhizodont expression) via signals from surrounding Materials ofthe Strepsodus were collectedat the Middle
tissues in their Paddock The Hawkins is thanked for target location. Again, this suggests site, Queensland. family to their and that the anterior access property making collecting work at the site position of the pectoral fin in osteo- possible through their invaluable support and Dr stracans hospitality. may have inhibited cranial mesodermal cells Tony and Guy Thulborn, and Angus and Tim Hamley (Queens- from differentiating into branchial mus- particular land Museum) collected the first rhizodont specimens. Help in cles, including the arch gill depressors, by instead the field and with preparation has been given by C. Burrow, T. 34 Z Johanson - Clavobranchialis musculature in fossil lungfishes
Colville, P. Crabb, B. Currie, R. Damiani, J. Ford, J, Garvey, Bemis WE, Lauder GV. 1986. Morphology and function ofthe
A. S. L. K. C. Hammerly, Howe, Masini, Parker, Northwood, feeding apparatus of the lungfish, Lepidosiren paradoxa
A. Walker and Work A. Yates. on the Ducabrook fossils is (Dipnoi). J. Morphol. 187: 81-108.
Australian Research Council supported by (ARC) grants no. Bishop IR, Foxon GEH. 1968. The mechanism of breathing in
A397009I5 and A00000629, and the author is funded an by the South American lungfish Lepidosirenparadoxa\ a radio- ARC Edwards invaluable Fellowship. Ashley provided assistance logical study. J. Zool. (London) 154: 263-272.
in the preparation of Neoceratodus and Jean Joss is material, - Campbell KSW, Barwick RE. 1987. Paleozoic lungfishes a
thanked for Univer- provided laboratory space (both Macquarie review. J. Morphol. Suppl. 1: 93-131.
also thank the Australian Museum for sity). I providing access Barwick RE. 1988a. Campbell KSW, Uranolophus: a reap- to vehicles and other resources. Ken Campbell, Peter Pridmore, praisal ofa primitive dipnoan. In: Jell, PA, ed. Devonian and The Royal Society London (Philosophical Transactions of the Carboniferous fish studies. Assoc. Australasian Palaeont.
The Western Australian Museum of Royal Society), (Records Mem. 7: 87-144.
the Western Australian Museum), Schweizerbart-Publishers Campbell KSW, Barwick RE. 1988b. Geological and palae- (PalaeontographicaA) and Francis & Taylor Publishers (Fossils ontological information and phylogenetic hypotheses. Geol.
and Strata) approved the use of ofthe images presented many Mag. 125: 207-227. this in paper. Finally, I want to thank Drs Sue Turner and Anne Campbell, KSW, Barwick RE. 1999, Dipnoan fishes from the Warren for allowing me to work on the Ducabrook material, Late Devonian Gogo Formation of Western Australia. Rec. and (he reviewers Profs K, S, W. Campbell and H. -P. Schultze West. Australian Mas. Suppl. 57: 107-138. for valuable comments. Campbell, KSW, Barwick RE. 2000. The braincase, mandible
and dental structures of the Early Devonian lungfish
Dipnorhynchus kurikae from Wee Jasper, New South Wales.
References Rec. Australian Mus. 52: 103-128,
Chang, M-M. 1991. Head exoskeleton and shoulder girdle of
Youngolepis. In; Chang M-M, Liu Y, Zhang G eds. Early Ahlbcrg PE, 1991. A re-examination of sarcopterygian inter- Vertebrates and Related Problems of Evolutionary Biology. relationships, with special reference to the Porolepiformes. Beijing; Science Press, 355-378. Zool. J. Linn. Soc. 103: 241-287.
Christ of B, Ordahl CP. 1995. Early stages chick somite de- Ahlbcrg PE, Johanson Z. 1997. The second tristichopterid velopment. Anat. Embryol. 191: 381-96. (Sarcopterygii, Osteolepiformes) from the Upper Devonian Daniel JF. 1934. The Elasmobranch Fishes. of Canowindra, New South Wales, Australia. J. Vert. Berkeley: Univer-
sity of California Press. Paleontol. 17: 653-673.
Dennis K, Miles RS. 1979a. A second Ahlberg PE, Johanson Z, Daeschler EB. 2001. The Late eubrachythoracid
arthrodire from Western Australia. Zool. Devonian kmgfish Soederberghia (Sarcopterygii; Dipnoi) Gogo, J. Linn. Soc.
67: 1-29. from Australia and North America, and its biogeographical
Dennis Miles RS. 19791). with implications. J. Vert. Paleontol. 21: 1-12. K, Eubrachythoracid arthrodires
Mechanics of the tubular rostral from Western Australia. Zool J. Alexander RMeN. 1969. feeding action ofa plates Gogo,
Linn. Soc. 67: 297-328. cyprinid fish. J. Zool (London) 159: 1-15.
Alexander RMcN. 1970. Mechanics of the feeding action of Dietrich S, Schubert FR, Healy C, Sharpe PT, Lumsden A.
various tclcost fishes. J. Zool. (London) 162: 145-156. 1998. Specification of the hypaxial musculature. Develop-
Andrews SM. 1973. Interrelationships of crossopterygians. In: ment 125: 2235-2249.
Brohmann Greenwood PH, Miles RS, Patterson C, eds. Interrelation- Dietrich S, Abou-Rebyeh F, H, Bladt F, Sonncn-
ships ofFishes. Zool. J. Linn. Soc. 53 Suppl 1: 137-177. berg-Riethmacher E,YaniaaiT,Lumsdcn A, Brand-Sabcri
Andrews SM. 1985. Rhizodont crossopterygian fish from the B, Birchmeier C. 1999. The role of SF/HGF and c-Met in
Dinantian of Foulden, Berwickshire, the ofskeletal Scotland, with a re- development muscle. Development 126: 1621-
evaluation of this group. Trans. R. Soc. Edinburgh: Earth 1629.
Sci. 76: 67-95. 1926. On Edgeworth EH. the developmentofthe cranial mus-
Andrews SM, Westoll, TS. 1970a. The postcranial skeleton of cles in Protopterus and Lepidosiren. Trans. R. Soc. Edin-
Eusthenopteron foordi Whiteaves. Trans. R. Soc. Edinburgh burgh 54: 719-734.
68: 207-329. KH. 1926. On ofthe Edgeworth the development cranial mus-
Andrews SM, TS. 19701). The skeleton of Westoll, postcranial cles in Protopterus and Lepidosiren. Trans. R. Soc. Edin- rhipidistians excluding Eusthenopteron Trans. R. Soc. Edin- burgh 54: 719-734. burgh 68: 391-489. Edgeworth FH. 1935. The Cranial Muscles ofVertebrates. Cam- Berwick RE, Campbell, KSW. 1996. A Late Devonian dipnoan, bridge: Cambridge University Press, Pillararhynchus, from Gogo, Western Australia, and its rela- Forey PJ. 1981. Rhabdoderma in the Carboniferous of the Brit- tionships. Palaeontographica A 239: 1-42. ish Isles. Palaeontology 24: 203-229. Bemis WE. 1987. Feeding systems of living Dipnoi: Anatomy Forey PJ. 1998. History of the Coelacanth Fishes. London: and Function. J. Morphol. Suppl. 1: 249-275. Chapman & Hall. Contributions to Zoology, 72 (1) — 2003 35
Fox RC, Campbell KSW, Barwick RE, Long JA. 1 995. A new Jarvik E. 1972. Middle and Upper Devonian Porolepiformes
osteolepiform fish from the Lower Carboniferous Raymond from East Greenland with special reference to Glyptolepis
Formation, Drummond Basin, Mem. Queens- Medd. Gron. 187: 1-307. , Queensland. n. sp, , groenlandica
land Mus. 38: 97-223. Jarvik E. 1980. Basic Structure and Evolution of Vertebrates,
GardinerBG. 1984. The relationships ofthe palaeoniscid fishes, Volume I. London: Academic Press.
a of review basedonnew specimens Mimia and Moythomasia Jeffery JE. 2001. Pectoral fins of rhizodontids and the evolu-
from Australia. Bull. the Upper Devonian of Western Bril. tion of pectoral appendages in the tetrapod stem-group.Biol.
Mus. (Nat. Hist.) Geol. 37: 173-428. J. Linn. Soc. 74: 217-236.
GV. 1994. Kinematics of in Gillis GB, Lauder feeding bluegill Jessen H. 1966. Die Crossopterygier des Oberen Plattenkalkes sunfish: distinction between is there a general aquatic cap- (Devon) der Bergisch-Gladbach-Paffrather Mulde (Rheini-
ture and transport behaviors? J. Exp. Biol. 198: 709-720. sches Schiefergebirge) unter Beriicksichtigung von ame-
DF. 1975. nouvel arthrodire du Goujet Dicksonosteus, un rickanischem und europaischem Onychodus-bAaXzn&X. Ark.
Devonian du Spitsberg; sur le squelette visceral remarques fur Zool. 18: 305-389. des Dolichothoraci. In: Lehman JP, ed. Problemes actuels do Jessen H. 1972. Schultergiirtel imd Pectoralflosse bei Actino- Evolution des Vertebras, Paris: Colloques Paleontologie: pterygiern. Fossils Strata 1: 1-101. Internationaux du Centre National de la Recherche Scientifique, Johanson Z. 1997. New Remigolepis (Placodermi; Antiarchi), 81-99. from Canowindra,NSW, Australia. Geol. Mag. 134:813-846. Goujet DF. 2001. Placoderms and basal gnathostome Johanson Z. in review. Placoderm hypobranchial muscles and apornorphies. In: Ahlberg PE, ed. Major Events in Early Ver- origins of the gnathostomejaw-loweringmusculature. J. Vert. tebrate Evolution: and Palaeontology, Phylogeny Develop- Paleo. ment. London: Systematics Association, 209-222. Johanson Z,Ahlbcrg PE. 1997. Newtristichopterid(Osteolepi- Gunther CA. 1871. of of Description Ceratodus, a genus ganoid formes: Sarcopterygii) from the Mandagery Sandstone fishes, recently discovered in rivers ofQueensland, Australia. Canowindra, N.S.W., Australia. (Famennian) near Trans. R. Trans. R. Soc. London 161: 511-571. Soc. Edinburgh Earth Sci. 88: 39-68, Hacker A, Guthrie, S. 1998. A distinct developmental pro- Johanson Z,Ahlberg PP. 1998. A complete primitive rhizodont gramme for the cranial paraxial mesoderm in the chick em- from Australia. Nature 394: 569-573. bryo. Development 125: 3461-3472. Johanson Z, Ahlberg PE. 2001. Devonian rhizodontids Haines L, Currie PD. 2001. Morphogenesis and evolution of (Sarcopterygii; Tetrapodomorpha) from East Gondwana. vertebrate appendicular muscle. J. Anat. 199: 205-209, Trans. R. Soc. Edinburgh Earth Sci. 92:43-74. HeintzA. 1932. The structure ofDinichthys contribution to our Johanson Z,Turner S, Warren A. 2000. First East Gondwanan knowledge of the Arthrodira. In: Gudger EW, ed. Archaic record ofSlrepsodus (Sarcopterygii; Rhizodontida)from the Fishes. Bashford Dean Memorial Volume, Article 4, 115- Lower CarboniferousDucabrook Formation, Central Queens- 224.
Geodiversitas 22: 161-169. Heintz land, Australia. A. 1934. Revision of the Estonian Arthrodira. Part I. branchial musculature and Jollic M. 1982. Ventral synapo- Family Homostiidae Jaekel, Publ. Geol. Inst. Univ. Tartu 38: questioned. Zool. J. Linn. Soc. 74: 35-47. 1-14, morphies
Hildebrand Kimmel CB, Miller CT & Keynes RJ. 2001. Neural crest M. 1974. Analysis of Vertebrate Structure. New patterning and the evolution of the jaw. J. Anat. 199: 105- York: John Wiley & Sons. Janvier 119. I’. 1980. Osteolepid remains from the Devonian of the 1979. mechanisms in teleosts Middle Lauder GV. Feeding primitive East, with particular reference to the endoskeletal shoulder and in the halecomorph fish Amia calva. J. Zool. 187: 543- girdle. In: Panchen AL, ed. The Terrestrial Envi- 578. ronment and the Origin of Land Vertebrates. Systematics Lauder GV. 1980. Evolution ofthe mechanism in Association Special Volume 15. London: Academic Press, feeding primi- 223-254. tive actinopterygian fishes: A functional anatomical analy-
Janvier sis of and Amia. J. 163; P. 1985. Les du Spitsberg.. Cahiers de Polyplerus, Lepisosteus Morphol. ... Cephalaspides, , D 283-317. Paleontologie (section Vertebres), Paris: Editions du Centre National de la Recherche Scientifique. Lauder GV. 1982. Patterns of evolution in the feeding mecha-
Janvier P. 1996. nism of fishes. Am. Zool. 22; 275-285. Early Vertebrates. Oxford Monographs on actinopterygian
Geology and Geophysics 33. Oxford: Oxford Science Publi- Lauder GV. 1983a. Functional design and evolution of the pha-
cations. ryngeal jaw apparatus in euteleostean fishes. Zool. J. Linn. Janvier P. 2001. Ostracoderms and the shaping of the Soc. IT: 1-38. gnathostome characters. In: Ahlberg PE, ed. Major Events Lauder GV. 1983b. Prey capture hydrodynamics in fishes: ex- in Vertebrate Early Evolution: Palaeontology,Phylogeny and perimental tests of two models. J. Exp. Biol. 104: 1-13.
Development. London: Systematics Association, 172-186. Lauder GV. 1985. Functional morphology ofthe feeding mecha- Janvier P, Marcoux J. 1976. Remarques sur la ceinture nism in lower vertebrates. In: Duncker HR, Fleischer G, eds.
scapulaire d’un choanate York: poisson Osteolepiforme des gres Functional Morphology ofVertebrates. New Springer devoniens rouges de I’Armutgozlek Tepe (Taurus lycien ori- Verlag, 210-289. ental, Turquie). Compt. rend. Acad. Sci. Paris 283: 619-622. Lauder GV, Liem KF. 1983. The evolution and interrelation- - 36 Z. Johanson Clavohranchialis musculature in fossil lungfishes
ships of the actinopterygian fishes. Bull. Mus. Comp. Zool. IVIotta PJ, Tricas TC, Hueter RE, Summers AP. 1997. Feed-
ISO: 95-197. ing mechanism and functional morphology of the jaws of
Lauder GV, Shaffer HB. 1985. Functional morphology ofthe the lemon shark Negaprion brevirostris (Chondnchthyes,
feeding mechanism in aquatic ambystomatid salamanders. Carcharhinidae). J. Exp. Biol. 200: 2765-2780.
J. Morphol. 185: 297-326. Motta PJ, Wilga CD. 1995. Anatomy ofthe feeding apparatus
Lebedev O. 1995. A new osteolepid fish from Russia. Bull. of the lemon shark (Megaprion brevirostris). J. Morphol.
Mus. nail Hist, natur. Paris 17: 287-341. 226:305-329.
Liem, KF. 1970. Comparative functional anatomy of the Molta PJ, Wilga CD. 1999. Anatomy ofthe feeding apparatus
Nandidiae (Pisces: Teleostei). Field. Zool. 56: 1-166. of the nurse shark fGinglymostoma cirratum). J. Morphol.
KF. 1980. of teleosts: 241: 33-60. Liem, Acquisition energy by adaptive
mechanisms and evolutionary patterns. In: Ali, MA, ed. En- Neidert AH, Virupannavar V, Hooker GW, Langeland JA.
vironmental ofFishes. Plenum Press: New 2001. Dlx and vertebrate evolution. Physiology York, Lamprey genes early
299-334. Proc. Natl. Acad. Sci. 98: 1665-1670.
Long JA 1987.A redescription ofthe lungfish Eoctenodus Hills Neyt C, Jagla K, Thisse C, Thisse, B, Haines L, Currie PD.
ofvertebrate 1929, with reassessment of other Australian records ofthe 2000, Evolutionary origins appendicular mus-
& Murchison 1828. Rec. West. cle. Nature 408: 82-86. genus Dipterus Sedgwick
Australian Mus. 13: 297-314, Norton SF, Brainerd E. 1993. Convergence in the feeding rhizodontiformfish mechanics of similar Long JA. 1989. A new from the Early Car- ecomorphologically species in the
boniferous of Victoria, Australia, with remarks on the Centrarchidae and Cichlidae. J. Exp. Biol. 176: 11-29.
of J. Vert. Paleonlol. 9: 1- Ortlahl LeDouarin NM. 1992. Two phylogenetic position the group. CP, myogenic lineages
17. within the developing somite. Development 114: 339-353.
Long ,IA. 1993. Cranial ribs in Devonian lungfishes and the Orvig T. 1975. Description, with special reference to the dermal
origin ofair-breathing. Assoc. Australasian Palaeontols. Mem. skeleton, of a new radotinid arthrodire from the Gedinnian of
15: 199-210. Arctic Canada. In: Lehman JP, ed. Problemes actuels de
Long JA, Barwick RE, Campbell KSW. 1997. Osteology and Paleontologie: Evolution des Vertebres. Paris; Colloques
functional morphology ofthe osteolepiform fish Gogonasus Internationaux du Centre National de la Recherche Scientifique,
andrewsi Long, 1985, from the Upper Devonian Gogo For- 41-72.
mation, Western Australia. Rec. West. Australian Mus. Suppl. Owen R. 1841 , Description ofthe Lepidosiren annectens. Trans.
53: 1-89. Linn. Soc. Land. 18: 327-361.
McMahon BR. 1969. A functional analysis of the aquatic and Pridmore PA, Campbell KSW, Barwick RE. 1994. Morphol-
aerial of African and ofthe respiratory movements an lungfish, ogy phylogenetic position holodipterandipnoans
Protopterus aethiopicus, with reference to the evolution of of the Upper Devonian Gogo Formation of northwestern
the lung-ventilation mechanism in vertebrates. J. Exp. Biol. Australia. Phil. Trans. R. Soc. London B 344: 105-164.
51: 407-430. Randall l)J. 1972. Respiration. In: Hardisty MW, Potter 1C,
Maiscy JG. 1989. Visceral skeleton and musculature ofa Late eds. The Biology ofLampreys. Volume 2. London: Academic
Devonian shark. J. Vert. Paleonlol. 9: 174-190. Press, 287-306.
Mallatt J. 1996. Ventilationand the origin ofjawed vertebrates: Reilly S. M. 1995. The ontogeny ofaquatic feeding behavior in
a new mouth. Zool. J. Linn. Soc. 117: 329-404. Salamandra salamandra: Stereotypy and isometry in feed-
198: Miles RS. 1971. The Holonematidae (placoderm fishes), a re- ing kinematics. J. Exp. Biol. 701-708.
view based on of The of kinematics new specimens Holonema from the Upper Reilly S. M. 1996. metamorphosis feeding in
Devonian of Western Australia. Phil. Trans. R. Soc. London Salamandrasalamandra and the evolution ofterrestrial feed-
263: 191-234. ing behavior. J. Exp. Biol. 199: 1219-1227.
Miles RS. 1977. Dipnoan (lungfish) skulls and the relationships Reilly SM, Lauder GV. 1990. The evolution of tetrapod feed-
ofthe from the Devonian Kinematic in Evo- group: a study based onnew species ing behaviour: homologies prey transport.
of Australia. Zool. J. Linn. Soc. 61: 1-328. lution 44: 1542-1557.
Miller CT, Schilling TF, Lee KH, Parker J, Kimmel CB. 2000. Roberts TDM. 1950. The respiratory movements of the lam-
sucker for encodes a zebrafish Endothelin-1 required ven- prey (Lampetrajluviatilis). Proc. R. Soc. Edinburgh (B) 64:
tral pharyngeal arch development. Development 127: 3815- 235-252. 3828. Romer AS, Parsons TS, 1986. The Vertebrate Body. Sixth Edi-
MiyakeT, McEachran J, Hall B. 1992. Edgeworth’s legacy of tion. Philadelphia: Saunders College Publishing.
cranial muscle with of muscles in Schaeffer B, KS. Reflections development an analysis Thomson, 1980. on agnathan-
the ventral gill arch region ofbatoid fishes (Chondrichthyes: gnathostome relationships. In: Jacobs LL, ed. Aspects of
Batoidea). J. Morph. 212: 213-246. Vertebrate History: Essays in Honor ofEdwin Harris Colbert.
Morrison SL, Campbell CK, Wright GM. 2000. Chondro- Flagstaff: Museum of Northern Arizona Press, 19-33.
genesis of the branchial skeleton in embryonic sea lamprey, SchillingTF, Kimmcl CB 1994. Segment and cell type lineage
Petromyzon marinus. Anal. Rec. 260: 252-67 restrictions during pharyngeal arch development in the
Moss SA. 1977. Feeding mechanisms in sharks. Am. Zool. 17: zebrafish embryo. Development 120: 483-494. 355-364. Schilling TF, Kimmel CB 1997. Musculoskeletalpatterning in Contributions to Zoology, 72 (1) - 2003 37
the pharyngeal segments ofthe zebrafish embryo. Develop- dipnoanspecies, Sorbitorhynchus deleaskitus, from the Lower
ment 124: 2945-2960. Devonian of Guangxi, China. Phil. Trans. R. Soc. London
Schultze H-P. 1975. Die Lungenfisch-Gattung Conchopoma (B) 340: 1-24.
(Pisces, Dipnoi). Senckenb. Lethaea 56: 191-251. Watson DMS, Gill EL. 1923. The structure ofcertain Paleozoic
Schultze H-P. 1977. Megapleuron zangerliA new dipnoanfrom Dipnoi. Zool. J. Linn. Soc. 35: 163-216.
the Pennsylvanian of Illinois. Fieldiana Geol. 33: 375-396. Wiley EO. 1979. Ventral gill arch muscles and the interrela- Schultze H-P. 1987. Dipnoans as sarcopterygians. 1 Morphol. tionships of gnathostomes, with a new classification of the
Suppl. 1: 39-74. Vertebrata. Zool. J. Linn. Soc. 67: 149-179.
Schultze H-P, Chorn J. 1998. Sarcopterygian and other fishes Wilga CD, Motta P. 1998. Feeding mechanism of the Atlantic
from the marine Upper Devonian of Colorado, U.S.A. Mitt. Guitarfish Rhinobatos lentiginosus: modulation ofkinematic
Mus. Naturkunde Berlin Geowiss. Reihe I: 53-72. and motor activity. J. Exp. Biol. 201: 3167-3184. Schultze H-l>, Cuumba SL. 2001, Diapilina and the characters Wilga CD, Wainwright PC, Motta P. 2000. Evolution of jaw
ofbasal actinopterygians. In: Ahlberg PE, ed. Major Events depression mechanics in aquatic vertebrates; insights from
in Early Vertebrate Evolution:Palaeontology,Phytogeny and Chondrichthyes. Zool. J. Linn. Soc. 71: 165-185.
Development. London: Systematics Association, 315-332. Young GC. 1980. A new Early Devonian placoderm from New
Schultze discussion HP, Marshall CR. 1993. Contrasting the use offunc- South Wales, Australia, with a of placoderm
tional complexes and isolated characters in lungfish evolu- phylogeny. Palaeontographica A 167: 1-76.
tion Assoc. Australasian Palaeontols. Mem. 15: 211-224. Zhu M, Schultze H-P. 2001. Interrelationships of basal Shaffer HB, Lauder GV. 1985. Patterns ofvariation in aquatic osteichthyans. In: Ahlberg PE, ed. Major Events in Early
ambystomatid salamanders: kinematics ofthe feeding mecha- Vertebrate Evolution: Palaeontology, Phylogeny and Devel- nism. Evolution 39: 83-92. opment. London: Systematics Association, 289-314. Stensio E. 1927. The Downtownian and Devonian Vertebrates Zhu M, Yu X, Janvier P. 1999. A primitive fossil fish sheds
of Spitsbergen. Ski: Svalbad Nordishavet 12: 1-391, light on the origin of bony fishes. Nature 397: 607-610. Stensio E. 1959. On the pectoral fin and shoulder girdle of the
arthrodires. K. Sv. Handl. 8: 1-229. Received: 15 2002 ... VetenskapsAkad. January
Wang S, Drapala V, Barwick RE, Campbell KSW. 1993. The