- Email: [email protected]
The ear region of Latimeria chalumnae: functional and evolutionary implications
Zoology 106 (2003): 233–242 © by Urban & Fischer Verlag http://www.urbanfischer.de/journals/zoology
The ear region of Latimeria chalumnae : functional and evolutionary implications Peter Bernstein Department of Systematic Zoology, Zoological Institute, University of Tübingen, Germany Received November 13, 2002 · Revised version received July 13, 2003 · Accepted September 8, 2003
Summary The anatomy of Latimeria chalumnae has figured prominently in discussions about tetrapod origins. While the gross anatomy of Latimeria is well documented, relatively little is known about its otic anatomy and ontogeny. To examine the inner ear and the otoccipital part of the cranium, a serial-sectioned juvenile coelacanth was studied in detail and a three-dimensional reconstruction was made. The ear of Latimeria shows a derived condition compared to other basal sarcopterygians in having a connection between left and right labyrinths. This canalis communicans is perilymphatic in nature and originates at the transition point of the saccule and the lagena deep in the inner ear, where a peculiar sense end organ can be found. In most gnathostomes the inner ears are clearly separated from each other. A connection occurs in some fishes, e.g. within the Ostariophysi. In the sarcopterygian lineage no connections between the inner ears are known except in the Actinistia. Some fossil actinistians show a posteriorly directed duct lying between the foramen magnum and the notochordal canal, similar to the condition in the ear of Latimeria, so this derived character complex probably developed early in actinistian history. Because some features of the inner ear of Latimeria have been described as having tetrapod affinities, the problem of hearing and the anatomy of the otical complex in the living coelacanth has been closely connected to the question of early tetrapod evolution. It was assumed in the past that the structure found in Latimeria could exemplify a transitional stage in otic evolution between the fishlike sarcopterygians and the first tetrapods in a functional or even phylogenetic way. Here the possibility is considered that the canalis communicans does not possess any auditory function but rather is involved in sensing pressure changes during movements involving the intracranial joint. Earlier hypotheses of a putative tympanic ear are refuted. Key words: Latimeria, inner ear, canalis communicans, Actinistia, Sarcopterygii, early tetrapods
Introduction Since the discovery of the living coelacanth in 1938, many aspects of its morphology have been intensively studied, so that anatomically Latimeria chalumnae is one of the best known fishes in the world today. Nevertheless, some morphological features are still poorly documented, especially those regarding very sensitive soft tissue complexes like the sense organs. Especially the investigation of the inner ear meets with particular difficulties, since its main part consists of very thin membranous canals and compartments, easily de-
stroyed or damaged by pressure changes when catching and pulling the animal from the deeper waters where Latimeria lives. Therefore, so far regrettably little research on the ear of Latimeria has been done. While the innervation pattern of the cranial nerves is known in great detail (Bemis and Northcutt, 1991; Northcutt and Bemis, 1993), only Fritzsch (1987a, b, 1992) examined the innervation of the end organs in the inner ear of Latimeria, reporting a basilar papilla-like structure at the transition point of the saccule and the lagena (Fritzsch, 1987a). These findings have been of systematic interest, since the presence of this putative
Corresponding author: Peter Bernstein, Department of Systematic Zoology, Zoological Institute, University of Tübingen, Auf der Morgenstelle 28, D-72076 Tübingen, Germany; phone: ++49-7071-29-74850, fax: ++49-7071-29-4634, e-mail: [email protected] 0944-2006/03/106/03-233 $ 15.00/0
P. Bernstein
basilar papilla has been used as an important character for defining a sister group relationship between the coelacanth fishes (Actinistia) and the Tetrapoda (Schultze, 1987; Marshall and Schultze, 1992; Zhu and Schultze, 2001), whereas now most scientists agree on a closer relationship between the Tetrapoda and the lungfishes (Dipnoi), using both morphological and molecular data (e.g. Rosen et al., 1981; Meyer and Wilson, 1990, 1991; Cloutier and Ahlberg, 1996; Zardoya and Meyer, 1996, 1997) as shown in Fig. 1. The inner ear of Sarcopterygii
The gross morphology of the inner ear of gnathostomes is quite uniform, in particular the upper part with the three semicircular canals and the utricle. Each of the three semicircular canals houses an ampulla with a sensory crista and is usually connected to the utriculus, which is more or less separated from the sacculus. The lower part with the sacculus and the lagena shows great variability within the Sarcopterygii, especially within the tetrapods. The reason for this complexity is mainly to be found in the evolution of a middle ear and the transformation of certain skeletal elements into ear ossicles, and, as a consequence thereof, the formation of special end organs in the inner ear along with a more elaborate perilymphatic system, functionally coupled with the new middle ear complex. In fishes no middle ear complex has ever evolved. The Weberian apparatus of the Ostariophysi in which modified vertebrae act as a kind of “ear ossicles” between the gas-filled swimbladder and the inner ear may be considered functionally analogous with the middle ear complex. But even without any such structures fishes
Fig. 1. Cladogram of the Sarcopterygii, showing recent crown groups and selected fossil forms (marked with an †). Based on data from Janvier (1996), Cloutier and Ahlberg (1996), Zhu and Schultze (2001) and Zhu and Yu (2002).
234
are not deaf but hear via the macula organs, which act as detectors for linear acceleration and are able primarily to detect passive movements caused in the near field of a sound source (Fritzsch, 2000). In many fishes or fish groups the macula organs do not equally participate in the task of hearing. In the ear of herrings, for example, the macula utriculi is the most important auditory receptor (Popper and Fay, 1993, 1997). Chondrichthyes seem to use the macula neglecta in conjunction with the macula sacculi for hearing (Corwin, 1981). Tetrapods, however, were confronted with a new set of problems when entering the terrestrial environment due to the different physical conditions in an air environment. They developed particular membranous end organs like the basilar papilla or amphibian papilla which in combination with an elaborate system of perilymphatic canals are responsible for the acute sense of hearing in many tetrapods, especially mammals. The middle ear acts as an impedance transducer, overcoming the problems of transferring airborne sound waves to the fluid medium of the inner ear, thus significantly increasing the sensibility of the ear. The evolution of tetrapod hearing
During the past 20 years our knowledge on the evolution of hearing in tetrapods has increased tremendously. Particular attention has been given to the evolutionary transformation of the otical region during the transition from an aquatic to a terrestrial lifestyle. The main interest was directed towards the anatomy of the ear in early tetrapods or members of the tetrapod stem group, since these are transitional forms leading from the fishlike ancestors to the fully terrestrial tetrapods. Many new fossils as well as new interpretations of old findings have revolutionized our knowledge of the vertebrate community at the end of the Devonian and the Early Carboniferous. Some of the fossils are so well preserved (e.g. Acanthostega, Clack, 1989, 1994, 1998) that the anatomy of the ear and related structures can be studied in great detail. So the stapes and its orientation in some very early tetrapods is now well known (Clack, 1983, 1992, 2002). Nevertheless, much remains unknown, because in fossils the soft parts are usually not preserved. This applies not only to the inner ear with its delicate membranous structures, but also to features of the middle ear region such as muscles, nerves, vessels and cartilage. Therefore, a closer look at the more recent tetrapods might be helpful in understanding the evolution of characters in the tetrapod otic region. Unfortunately, this approach has its own limitations. Most living tetrapods have very specialized ears and are poor models for early otic evolution. So, Clack (1998) is surely right in saying that “returning to the earliest tetrapods, I am left with the Zoology 106 (2003) 3
The ear region of Latimeria
overwhelming impression that we must look to fishes such as basal actinopterygians, lungfishes and coelacanths for the nearest analogue of their hearing abilities and inner ear development.” Looking at the living coelacanth, this is all the more important, since some kind of functional tympanic ear was proposed in this taxon (Fritzsch, 1992). The investigation of the auditory function in fishes usually requires electrophysiological, behavioral or even more advanced methods like the auditory brainstem response technique (Kenyon et al., 1998). Because of the elusive lifestyle and the protected status of Latimeria, it seems very unlikely that the coelacanth could be examined in such a way in the foreseeable future. For this reason, we have to check whether the anatomy may provide further insight into the problems mentioned above.
otic capsule is not completely enclosed by cartilage and often lacks a medial wall to the cranial cavity, the inner ear of Latimeria is well encapsulated. The anterior border is situated at the origin of the anteriorly directed processes of the prootic, these being in close contact to the connecting processes of the basisphenoid as part of the intracranial joint. The braincase of Latimeria is divided into an anterior ethmosphenoidal and a posterior otoccipital part. These parts are connected via the intracranial joint, a feature once common in early vertebrates but now restricted to the living coelacanth. All characters involved in hearing are situated in the otoccipital part, which houses the otic capsule and most of the brain.
Materials and methods A juvenile specimen of Latimeria chalumnae Smith, 1939 (CCC-No. 162 K, total length = 310 mm) served as the main object of examination. The specimen had been originally deep-frozen and then preserved in formalin. The state of preservation was good but not perfect. After embedding in celloidin, serial transverse sections 40 µm thick were cut. These were stained with Azan according to Heidenhain. Besides serving as a basis for conventional light microscopic investigation, the serial sections were used to make a three-dimensional reconstruction of the otoccipital part of the cranium using a modified wax plate technique (Born, 1883). Instead of wax, polystyrene was utilized, which is much easier to use. An adult specimen of Latimeria chalumnae Smith, 1939 (CCC-No. 161, SZ 10064), preserved in formalin, was used for estimating the morphometric differences between the juvenile and adult conditions. Both specimens are from the collection of the Department of Systematic Zoology at the University of Tübingen. The specimen is one of 26 embryos that were still inside the mother’s belly at the time of catch, ranging in length from 308 to 358 mm. As far as could be estimated, they were almost ready to be born. The overall proportions are very similar to the adult ones, though a slight lengthening of the skull and the body occurs during later ontogeny.
Results The otic capsule
The otic capsule of Latimeria lies entirely in the otoccipital part of the cranium (Fig. 2 –3). While in many basal actinopterygians and especially in lungfishes the Zoology 106 (2003) 3
Fig. 2. Three-dimensional model of the otoccipital part of a juvenile Latimeria. Parts of the tandem articulation are also shown. Left lateral view. Membrane bones are omitted. The sectional planes of some of the following illustrations are indicated by the respective figure numbers. The region of the thin “tympanumlike” skin is highlighted with a red circle. ah, anterior head of hyomandibula; cartM, Meckelian cartilage; crp, crista parotica; hyom, hyomandibula; inhy, interhyal; opcar, opercular cartilage; ph, posterior head of hyomandibula; pq, palatoquadrate; pro, connecting processes of prootic; qd, quadrate; symp, symplectic; styl, stylohyal.
235
P. Bernstein
Fig. 3. Skull of an adult Latimeria chalumnae. Left lateral view. Redrawn from Millot et al. (1978). Abbreviations see Fig. 2.
Fig. 5. Transverse section through the skull of a juvenile Latimeria chalumnae. On the right hand side, the foremost part of the spiracular pouch lies between the connecting processes of the prootic and the expansive palatoquadrate. The thin skin between the spiracular pouch and the outer medium is marked with a red circle. bm, basicranial muscle; br, brain; bsph, basisphenoid; eptg, entopterygoid; icav, intracranial cavity; itmp, intertemporal; mcav, mouth cavity; mptg, metapterygoid; nc, notochord; pq, palatoquadrate; pro, connecting processes of prootic; spc, spiracular chamber.
Fig. 4. Transverse section through the skull of a juvenile Latimeria chalumnae. The otic capsule is completely enclosed by cartilage except for the medial foramen of the canalis communicans, which originates at the end organ. bm, basicranial muscle; br, brain; cc, canalis communicans; crc, crus commune; eo, end organ; for, foramen in medial wall of otic capsule; icav, intracranial cavity; mcav, mouth cavity; nc, notochord; ophy, opercular process of hyomandibula; stmp, supratemporal.
The dorsolateral wall of the otic capsule is well developed, particularly at the hyomandibular insertion point. Here, a prominent lateral crest (crista parotica) or otic shelf strengthens the articulation area. Medially, the inner ear is separated from the notochord and the cranial cavity by a partially ossified medial wall, which is perforated by foramina for nerves and other elements 236
such as the endolymphatic duct. Ossification of the medial wall is to be seen close to the notochord only, obviously reinforcing the otic capsule against forces active during movements within the intracranial joint. The same applies to the floor of the capsule at the inserting region of the basicranial muscle, where the skeletal elements have to withstand strong forces during contraction of this muscle (Fig. 4). The spiracular pouch and the hyomandibula
Latimeria has a spacious spiracular pouch but no spiracular opening. At the level of the intracranial joint, where the upper portion of the palatoquadrate lies, the anterior end of the spiracular pouch is dorsally separated from the outside by a thin membrane-like skin (Fig. 2, 5). It extends backwards and is in broad contact with the mouth cavity at the level where the prootic processes connect to the otoccipital roof. The posterior margin extends to the space above the lateral crest where it reaches the anterior articulation of the hyomandibula. Zoology 106 (2003) 3
The ear region of Latimeria
The hyomandibula has four articulations with other skeletal elements. Two articulations connect the hyomandibula to the braincase, a posteriorly directed process articulates to the opercular, and a ventral process contacts the interhyal. The anterior, more ventral articulation to the otic capsule is associated with strong ligamentous connective tissue, whereas the posterior, more dorsal insertion shows weaker ligaments. The inner ear
Fig. 6. The right inner ear of an adult Latimeria chalumnae in medial view. From Millot and Anthony (1965), modified. Note the dissected canalis communicans. The position of the end organ is marked in red. cc, canalis communicans; dend, endolymphatic duct; eo, end organ; lag, lagenar recess; sac, sacculus. VIII, cranial nerve 8.
The inner ear of Latimeria was described in detail by Millot and Anthony (1965). The upper part shows the usual gnathostome pattern with a moderate-sized utricle and three perpendicularly oriented semicircular canals (Fig. 6). In the lower part the sacculus is very prominent. It houses a very large otolith on the medial side and is connected to a posteriorly directed and separated lagenar recess. At the transition point of the saccule and the lagenar recess an orifice is situated, marking the origin of a bilateral connection between the inner ears, known as canalis communicans. Its exact origin is situated where a peculiar membrane-like structure can be found deeper in the inner ear (Fig. 4, 7– 8). From there the canal passes medially through a foramen in the medial wall of the otic capsule and extends backwards between the notochord and the otic capsule. For most of its course the canalis communicans is completely enclosed by cartilage (Fig. 9).
Fig. 7. The canalis communicans and the end organ in the inner ear of Latimeria chalumnae. This figure shows an enlarged view of a similar plane as Fig. 4 in histological display. cc, canalis communicans; eo, end organ; for, foramen in medial wall of otic capsule; nc, notochord; sac, sacculus.
Fig. 8. A further enlarged view shows the peculiar end organ in the inner ear of Latimeria chalumnae lying in the endolymphatic space. Above, the origin of the canalis communicans is situated. cc, canalis communicans; eo, end organ; es, endolymphatic space; ps, perilymphatic space.
Zoology 106 (2003) 3
237
P. Bernstein
canals disappear not far from that place, both parts of the canalis communicans pass further backwards, eventually merging in the occipital region between the foramen magnum and the notochord (Fig. 9D). Where the canalis communicans enters the saccular/lagenar region, an apparent end organ with a membrane spans the orifice between the endo-/perilymphatic spaces (Fig. 7– 8). Clearly, this is the structure regarded by Fritzsch (1987a) as a basilar papilla. It seems indeed to be an innervated end organ, consisting of a membrane and possibly an overlying group of cells (Fig. 8). Its size is about 1 mm and it seems to span the entire distance between the canalis communicans and the saccular/lagenar orifice. It is apparently innervated by fibers from the lagenar branch of the statoacustic nerve VIII. There are two completely separated spaces, one endolymphatic and one perilymphatic, with the end organ lying in the endolymphatic portion (Fig. 8). The perilymphatic space is enclosed by a duct (shown collapsed in Fig. 7) which traverses the medial wall of the otic capsule and passes further backwards without forming an expanded pouch as depicted by Fritzsch (1992). Instead, the foramen in the medial wall is comparatively narrow with the duct passing immediately posteriorly after traversing this foramen (Fig. 4).
Discussion A tympanic ear in Latimeria?
Fig. 9. Four consecutive serial sections show the passage of the canalis communicans in posterior direction. The connection of the lower and upper canals is marked by an arrow. cc, canalis communicans; icav, intracranial cavity; nc, notochord; ucc, upper part of canalis communicans.
Not far behind the foramen for the canalis communicans another set of canals originates at the floor of the cranial cavity, dorsally of the canalis communicans (Fig. 9A). These canals become enclosed by cartilage on their passage backwards, so that in cross-sections a double set of canals can be seen for quite a distance (Fig. 9B–D). Near the posterior end of the otic capsule the upper canals and the canalis communicans are connected by a very tiny duct on each side (Fig. 9C). While the upper 238
The identification of a papilla-like receptor end organ in the inner ear of Latimeria and the finding of a thin membrane-like skin between the spiracular pouch and the outer medium led Fritzsch (1992) to hypothesize a tympanic function of the auditory system in the coelacanth. The spiracular pouch would in this case act as a pressure-detection system with the thin skin functioning as a vibrating membrane much as the tympanum in tetrapods. Furthermore, Fritzsch (1992) proposed several ways how pressure changes could be transmitted to the receptor end organ he interpreted as a basilar papilla. These include possible pressure differences between the spiracular pouch and the intracranial cavity, pressure differences between the spiracular pouch and the belly transmitted by the canalis communicans, participation of the lower jaw and the hyomandibula in inducing pressure differences between the inner ear and the brain cavity and movements within the intracranial joint. To evaluate these hypotheses one has to look at the function of the intracranial joint and the hyomandibula, the anatomy of the inner ear and the canalis communicans and the relationships among those structures. In order to act as a resonance chamber, the spiracular Zoology 106 (2003) 3
The ear region of Latimeria
pouch of Latimeria would have to be filled with gas. The mere existence of a gas-filled chamber like the swimbladder somewhere in the body improves the hearing abilities in fish (Blaxter, 1981), and the closer this gas-filled chamber is to the ear the better. In anabantoid fishes (Schneider, 1941; Yan, 1998) the improvement of hearing capabilities by air-filled chambers near the otic capsules is well known. Fishes in other taxonomic groups have developed other means of auditory enhancement. These include the clupeids, with otic bullae connected to the swimbladder by small tubes, and the Ostariophysi (e.g. carp, catfishes), with a series of bony elements between the swimbladder and the ear, known as the Weberian apparatus. These bony elements act as “ear ossicles”, transmitting sound-generated vibrations from the swimbladder to the inner ear. The superficial resemblance of the inner ear connection in the Ostariophysi and Latimeria was already noticed by Millot and Anthony (1965), but they also pointed out the lack of any structures transmitting pressure waves to the ear of the coelacanth and the impossibility of the coelacanth’s fat-filled swimbladder to act as a resonance chamber. Moreover, the swimbladder is not situated in the vicinity of the ear. There is no obvious sound-transmitting apparatus from the abdominal region to the canalis communicans. Fritzsch (1992) pointed out the discovery of some fossil actinistians with an apparently ossified swimbladder and its resemblance to the ossified swimbladders of some recent ground-dwelling catfishes (Chranilov, 1929; Fiedler, 1991). Indeed, these catfishes may use their swimbladder as a resonance chamber for hearing, but one has to keep in mind that these fishes are ostariophysines and possess the Weberian apparatus, which is completely lacking in Latimeria. There is experimental evidence that these catfishes with an ossified swimbladder-wall even have decreased hearing ability (Ladich, 1999), so the discovery of ossified swimbladders in fossil actinistians probably does not indicate a pathway for pressure waves from the swimbladder to the inner ear, rather the opposite. The ossification of the swimbladder probably is an adaptation in ground-dwelling or deep-sea fishes to better withstand the high water pressure. Since gas is easily compressed, it would be hard for any fish living at greater depths to maintain the gas filling in its swimbladder. The same might be true for Latimeria, where the swimbladder is fat-filled. The fatty substance gives buoyancy and is not easily compressed in depths of 150 –700 m, where Latimeria usually lives (Fricke and Plante, 1988). Furthermore, the existence of an ossified swimbladder or lung in the Actinistia seems to be restricted to Mesozoic forms like Undina or Macropoma (Janvier, 1996). Since the Devonian genus Diplocercides already shows a canalis communicans (Fig. 10) but no ossified swimZoology 106 (2003) 3
bladder (Jarvik, 1980), a functional correlation between these two characters does not seem very likely. Another problem regarding a possible gas-filled spiracular pouch is the great depth Latimeria lives in. As noted above, it is not easy to maintain a gas-filled cavity against the high water pressure in deep water. And how would this cavity be filled with gas? Only two explanations would be reasonable: either Latimeria has to take up air at the surface or there has to be a gas-secreting gland in the spiracular pouch. While there is no evidence for the latter possibility, the first is not supported by observation either. The depth Latimeria lives in (ca. 200 m and deeper) does not support such a notion. The role of the hyomandibula
The use of a gas-filled spiracular chamber near the ear capsule and the hyomandibula as a receptor of far-field sound in fishes has been discussed by van Bergeijk (1966). The proposed close functional connection between these elements is in line with the well established homologies between the hyomandibula and the stapes on the one hand and between the spiracular pouch and the tetrapod middle ear cavity on the other hand. In the traditional view of otic evolution during the transition from water to land, the position of the hyomandibula close to the ear capsule and its vicinity to the spiracular pouch made for an easy transformation of these elements into parts of an air-adapted tetrapod ear (Lombard and Bolt, 1979). Although this view has been challenged by the discovery of bulky stapes in the earliest tetrapods, which were seemingly quite inappropriate for aerial hearing and suitable for sensing ground vibra-
Fig. 10. Reconstructed brain case of the Devonian actinistian Diplocercides, occipital view. The canalis communicans is situated in the same position as in Latimeria, between the foramen magnum and the notochordal canal. Redrawn from Jarvik (1980). cc, canalis communicans; fm, foramen magnum; nc, notochord.
239
P. Bernstein
tions only (Clack, 1992), the possibility remains that the condition of these structures in Latimeria could present a transitory stage in a functional, or even phylogenetic context, if the conditions in Latimeria are interpreted as ancestral as proposed by Fritzsch (1992). According to Eaton (1939), in the early sarcopterygians the hyomandibula originally had five articulations with other parts of the skull: two heads inserting at the otic capsule, an opercular process, an articulation with the quadrate and another articulation with the ceratohyal. If we accept this as the sarcopterygian groundplan, Latimeria has four of these articulations and has lost the one to the quadrate, since the quadrate-Meckelian cartilage joint is separated from the symplectic-Meckelian cartilage joint due to the peculiar tandem articulation of the coelacanth’s jaws. While this constellation is surely derived, the hyomandibula of Latimeria is part of a complex skull kinesis and certainly has an important function in integrating the movements of the various skull elements, probably just as in the extinct “osteolepiforms” (Thomson, 1967). Only after the hyomandibula lost this function, it could be used in a different functional context and got involved in hearing (Thomson, 1966b). Although there are quite a few incompatible theories on the skull kinetics of Latimeria (e.g. Thomson, 1966a; Robineau and Anthony, 1973; Alexander, 1973; Adamicka and Ahnelt, 1976; Lauder, 1980), it seems to be clear that the hyomandibula of Latimeria is in no way “unfunctional”, but participates in the complicated sequence of movements during mouth opening and closing, prey capture and breathing. The strong attachment of the hyomandibula to the otic capsule, allowing only limited mobility, was already noted by Alexander (1973). The existence of a robust ligamentous connection of the anterior hyomandibular head articulating to the otic capsule seems to support this view. A hingelike movement of the hyomandibula against the braincase is all that seems to be possible. The lateral wall of the otic capsule where the medial heads of the hyomandibula insert is especially thick, apparently to strengthen this critical and strained location. The perilymphatic space connected to the end organ in the inner ear is on the opposite side near the medial wall of the otic capsule. A functional connection between the perilymphatic canalis communicans and the hyomandibula seems very unlikely. A similar situation can be found regarding a possible connection of the spiracular pouch and the hyomandibula. The latter is far away from the end organ or the canalis communicans and only in peripheral contact with the posteriormost offshoot of the spiracular pouch. The obvious conclusion is that there is no functional connection between the hyomandibular apparatus and the inner ear end organs or the canalis communicans. 240
The canalis communicans and the “basilar papilla”
The anatomy and the location of the mentioned elements do not support a functional relationship (at least not an auditory one) between those elements and the canalis communicans and the peculiar end organ in the inner ear. There has to be another function of this inner ear connection, which has obviously been present since the early actinistian evolution, as the skeletal canalis communicans seemed to be already present in the Devonian actinistian Diplocercides (Jarvik, 1980). As far as can be guessed from the description of the reconstructed brain case, Diplocercides had a canalis communicans very similar to the coelacanth’s one (Fig. 10). Unfortunately, this canal is not found in other fossil actinistians. That does not necessarily mean that there was none, because in most fossil actinistians the region surrounding the canalis communicans was obviously cartilaginous and therefore not preserved (Forey, 1998). The canalis communicans originates at the observed end organ deep in the inner ear and reaches the foramen in the medial wall of the otic capsule via a duct and not via a perilymphatic sac as proposed by Fritzsch (1992). There is no connection between this duct and the undifferentiated perilymphatic space surrounding the sacculus. The main connection to the intracranial space is not at the place of the perilymphatic foramen in the medial wall but via tiny ducts leading to the upper set of canals paralleling the canalis communicans proper (Fig. 9). Those ducts are connected to the intracranial space and are surely able to transmit any pressure changes inside the intracranial cavity to the inner ear. The possibility of transmission of pressure changes via this pathway during movements within the intracranial joint was recognized by Fritzsch (1992), but in this case an auditory function of the whole complex must be questioned. For any auditory specialization (pressure detection and transduction of displacement) there has to be a low-density pathway between the hair cells of the end organ and a displacement transducer via a periotic fluid system as found in the tetrapod ear and the ears of some specialized teleosts (Lombard and Hetherington, 1993), which could not be found in the living coelacanth. There is no evidence for any participation of the intracranial joint in hearing. If the end organ is sensitive to pressure changes within the intracranial cavity, it might be part of controlling the complex pattern of movements involved in the jaw mechanics of Latimeria. In this case it would be rather a specialized organ possibly unique to the actinistian lineage and not homologous to the basilar papilla in tetrapods. Taking into account the points mentioned above, we have to conclude that the ear of Latimeria is hardly suitable as a functional transitional form between the ear of early fishlike sarcopterygians and the modern tetrapod ear. Zoology 106 (2003) 3
The ear region of Latimeria
These results seem to show once more the complexity of the character mosaic of Latimeria that every systematist is confronted with. While some characters of the coelacanth apparently are plesiomorphic such as the structure of the brain (Millot and Anthony, 1965), the secondary capillary system in the gills (Vogel et al., 1998) or the myoseptal architecture of the trunk muscles (Gemballa and Ebmeyer, 2003), others are clearly autapomorphic to Latimeria or actinistians as a group, like the rostral organ (Bemis and Hetherington, 1982) or the unique tandem articulation of the jaws (Adamicka and Ahnelt, 1976). Concerning the inner ear of Latimeria, the little research previously done indicates a plesiomorphic condition of the hair cell pattern in the macula lagenae, without any affinities to the tetrapod state (Platt, 1994). The end organ at the origin of the canalis communicans may not be similar to the basilar papillae found in basal tetrapods such as salamanders or plesiomorphic anurans. In the ears of these forms, the basilar papillae as well as the amphibian papillae are easily identified by their location and appearance, with the hair cells conspicuously distinguishable. The end organ of Latimeria does not easily allow such a homologization and is morphologically different from the papillae of lissamphibians. Fritzsch (1987, 1992) is right in pointing out the similarity of the end organ’s innervation by a twig of the lagenar branch of the VIIIth nerve to the innervation pattern of the basilar papilla in tetrapods. But does that necessarily mean that both structures are homologous? We have evidence that the sensory epithelia in the vertebrate ears developed by the splitting of a single sensory precursor (Fritzsch et al., 2002). If so, the end organ in the ear of Latimeria could be the part representing the homologon to the basilar papilla in tetrapods, but because there is also evidence for the independent origin of the lagena (Fritzsch, 1992; Fritzsch et al., 2002) this could likewise point to an independent origin of the end organ coupled to the lagena. This would be the basilar papilla in tetrapods but could be another end organ in other vertebrate groups, e.g. the coelacanth. The basilar papilla in tetrapods is coupled to an auditory function; using the term “basilar papilla” for an end organ found in the inner ear of a vertebrate could easily tempt one to assume an auditory function similar to the ear of tetrapods. As long as we have no evidence for such a function of the end organ in the coelacanth’s ear, I would recommend to avoid the term “basilar papilla”. In a phylogenetic sense the results support the uniqueness of the coelacanth’s cranial functional morphology, even more so if the end organ and the canalis communicans are involved in controlling the mechanics of the intracranial joint rather than in hearing. The condition Zoology 106 (2003) 3
of the Devonian coelacanth Diplocercides probably shows us the early evolution of this character. Therefore, the end organ in the inner ear of Latimeria may not be represented as synapomorphic with the tetrapods. For this reason I prefer the phylogeny of living sarcopterygians as shown in Fig. 1, showing the dipnoi as closely related to the tetrapods, with the actinistians as the sister group of the clade formed by the two other groups.
Acknowledgements I would like to thank Prof. Dr. W. Maier for providing access to the material of Latimeria and the permission to use the facilities at the Lehrstuhl für Spezielle Zoologie at the University of Tübingen. The specimens of Latimeria were generously provided by Prof. Dr. H. Fricke. Many helpful discussions with S. Gemballa, M. Hoffmann and M. Sánchez-Villagra provided invaluable advice. M. Sánchez-Villagra read an earlier draft of this paper and gave many useful comments. I also want to thank an anonymous reviewer and Bernd Fritzsch for many helpful remarks. K. Hagen was very helpful in translating parts of the work of Millot and Anthony (1965) from French into German. M. Meinert, T.-T. Fußnegger and G. Schmid provided technical assistance. This study was supported by a grant from the Landesgraduiertenförderung Baden-Württemberg.
References Adamicka, P. and H. Ahnelt. 1976. Beiträge zur funktionellen Analyse und zur Morphologie des Kopfes von Latimeria chalumnae Smith. Ann. Naturhist. Mus. Wien 80: 251–271. Alexander, R.M. 1973. Jaw mechanisms of the Coelacanth Latimeria. Copeia 1: 156 –158. Bemis, W.E. and T.E. Hetherington. 1982. The rostral organ of Latimeria chalumnae: morphological evidence of an electroreceptive function. Copeia 2: 467– 471. Bemis, W.E. and R.G. Northcutt. 1991. Innervation of the basicranial muscle of Latimeria chalumnae. Environ. Biol. Fish. 32: 147–158. Bergeijk, W.A. van. 1966. Evolution of the sense of hearing. Amer. Zool. 6: 371–377. Blaxter, J.H. S. 1981. The swimbladder and hearing. In: Hearing and Sound Communication in Fishes (W. N. Tavolga, A.N. Popper and R.R. Fay, eds.). Springer, New York, pp. 61–71. Born, G. 1883. Die Plattenmodellirmethode. Arch. Mikros. Anat. 22: 584–599. Clack, J.A. 1983. The stapes of the Coal Measures embolomere Pholiderpeton scutigerum Huxley (Amphibia: Anthracosauria) and otic evolution in early tetrapods. Zool. J. Linn. Soc. 79: 121–148. Clack, J.A. 1989. Discovery of the earliest-known tetrapod stapes. Nature 342: 425–427. Clack, J.A. 1992. The stapes of Acanthostega gunnari and the role of the stapes in early tetrapods. In: The Evolutionary Biol-
241
P. Bernstein
ogy of Hearing (D. B.Webster, R. R. Fay and A. N. Popper, eds.). Springer, New York, pp. 405 – 420. Clack, J.A. 1994. Earliest known tetrapod braincase and the evolution of the stapes and fenestra ovalis. Nature 369: 392–394. Clack, J.A. 1998. The neurocranium of Acanthostega gunnari Jarvik and the evolution of the otic region in tetrapods. Zool. J. Linn. Soc. 122: 61– 97. Clack, J.A. 2002. Gaining Ground: The Origin and Evolution of Tetrapods. Indiana University Press, Bloomington. Cloutier, R. and P.E. Ahlberg. 1996. Morphology, characters, and the interrelationships of basal sarcopterygians. In: Interrelationships of Fishes (M.L.J. Stiassny, L.R. Parenti and G.D. Johnson, eds.). Academic Press, San Diego, pp. 445 –479. Corwin, J.T. 1981. Audition in Elasmobranchs. In: Hearing and Sound Communication in Fishes (W.N. Tavolga, A.N. Popper and R.R. Fay, eds.). Springer, New York, Berlin, Heidelberg, pp. 81–105. Eaton, T.H. 1939. The crossopterygian hyomandibular and the tetrapod stapes. J. Wash. Acad. Sci. 29: 109 –117 Fiedler, K. 1991. Lehrbuch der speziellen Zoologie. Band 2: Wirbeltiere, Teil 2: Fische. Fischer, Stuttgart. Forey, P.L. 1998. History of the Coelacanth Fishes. Chapman & Hall, London. Fricke, H., O. Reinicke, H. Hofer and W. Nachtigall. 1987. Locomotion of the coelacanth Latimeria chalumnae in its natural environment. Nature 329: 331– 333. Fricke, H. and R. Plante. 1988. Habitat requirements of the living coelacanth Latimeria chalumnae at Grande Comore, Indian Ocean. Naturwiss. 75: 149 –151. Fritzsch, B. 1987a. Inner ear of the coelacanth fish Latimeria has tetrapod affinities. Nature 327: 153 –154. Fritzsch, B. 1987b. Die Papilla basilaris von Latimeria chalumnae und die Evolution von amphibientypischen auditiven Sinnesepithelien. Verh. Dtsch. Zool. Ges. 80: 145 –146. Fritzsch, B. 1992. The water-to-land transition: evolution of the tetrapod basilar papilla, middle ear, and auditory nuclei. In: The Evolutionary Biology of Hearing (D.B. Webster, R.R. Fay and A.N. Popper, eds.). Springer, New York, pp. 351–375. Fritzsch, B. 2000. Microscopic functional anatomy: Sensory systems: Hearing. In: The Laboratory Fish (G.C. Ostrander, ed.). Academic Press, London, pp. 480 – 487. Fritzsch, B., K.W. Beisel, K. Jones, I. Fariñas, A. Maklad, J. Lee and L.F. Reichardt. 2002. Development and evolution of inner ear sensory epithelia and their innervation. Journal of Neurobiology 53: 143 –156. Gemballa, S. and L. Ebmeyer. 2003. Myoseptal architecture of sarcopterygian fishes and salamanders with special reference to Ambystoma mexicanum. Zoology 106: 29 – 41. Janvier, P. 1996. Early Vertebrates. Clarendon Press, Oxford. Jarvik, E. 1980. Basic Structure and Evolution of Vertebrates. Academic Press, London. Kenyon, T.N., F. Ladich and H.Y. Yan. 1998. A comparative study of hearing ability in fishes: the auditory brainstem response approach. J. Comp. Physiol. A 182: 307– 318. Ladich, F. 1999. Did auditory sensitivity and vocalization evolve independently in otophysan fishes? Brain Behav. Evol. 53: 288 – 304. Lauder, G.V. 1980. The role of the hyoid apparatus in the feeding mechanism of the Coelacanth Latimeria chalumnae. Copeia 1: 1– 9. Lombard, R.E. and J.R. Bolt. 1979. Evolution of the tetrapod ear: an analysis and reinterpretation. Biol. J. Linn. Soc. 11: 19–76. Lombard, R.E. and E.H. Hetherington. 1993. Structural basis of hearing and sound transmission. In: The Vertebrate Skull, Volume 3: Functional and Evolutionary Mechanisms (J. Hanken
242
and B.K. Hall, eds.). University of Chicago Press, Chicago, pp. 241–302. Marshall, C. and H.-P. Schultze. 1992. Relative importance of molecular, neontological, and paleontological data in understanding the biology of the vertebrate invasion of the land. J. Mol. Evol. 35: 93–101. Millot, J. and J. Anthony. 1965. Anatomie de Latimeria chalumnae. II. Système nerveux et organes des sens. C.N.R.S., Paris. Millot, J., J. Anthony and D. Robineau. 1978. Anatomie de Latimeria chalumnae. III. Appareil digestif, appareil respiratoire, appareil uro-génital, glandes endocrines, appareil circulatoire, téguments, écailles, conclusions générales. C.N.R.S., Paris. Meyer, A. and A.C. Wilson. 1990. Origin of tetrapods inferred from their mitochondrial DNA affiliation to lungfish. J. Mol. Evol. 31: 359–364. Meyer, A. and A.C. Wilson. 1991. Coelacanth’s relationships. Nature 353: 219. Northcutt, R.G. and W.E. Bemis. 1993. Cranial nerves of the coelacanth, Latimeria chalumnae (Osteichthyes: Sarcopterygii: Actinistia), and comparisons with other Craniata. Brain Behav. Evol. 42 [Suppl. 1]: 1–74. Platt, C. 1994. Hair cells in the lagenar otolith organ of the coelacanth are unlike those in amphibians. J. Morph. 220: 381. Popper, A.N. and R.R. Fay. 1993. Sound detection and processing by fish: critical review and major research questions. Brain Behav. Evol. 41: 14 –38. Popper, A.N. and R.R. Fay. 1997. Evolution of the ear and hearing: issues and questions. Brain. Behav. Evol. 50: 213–221. Robineau, D. and J. Anthony. 1973. Biomécanique du crâne de Latimeria chalumnae (Poisson crossoptérygien coelacanthidé). C.R. Acad. Sci. Paris Sér. D 276: 1305–1308. Rosen, D.E., P.L. Forey, B.G. Gardiner and C. Patterson. 1981. Lungfishes, tetrapods, paleontology, and plesiomorphy. Bull. Am. Mus. Nat. Hist. 167: 159 –276. Schneider, H. 1941. Die Bedeutung der Atemhöhle der Labyrinthfische für ihr Hörvermögen. Z. vergl. Physiol. 29: 172 –194. Schultze, H.-P. 1987. Dipnoans as Sarcopterygians. J. Morphol. Suppl. 1: 39 –74. Thomson, K.S. 1966a. Intracranial mobility in the coelacanth. Science 153: 999 –1000. Thomson, K.S. 1966b. The evolution of the tetrapod middle ear in the rhipidistian-amphibian transition. Amer. Zool. 6: 379 –397. Thomson, K.S. 1967. Mechanisms of intracranial kinetics in fossil rhipidistian fishes (Crossopterygii) and their relatives. Zool. J. Linn. Soc. 46: 223–253. Vogel, W.O.P., G.M. Hughes and U. Mattheus. 1998. Non-respiratory blood vessels in Latimeria gill filaments. Phil. Trans. R. Soc. Lond. B 353: 465– 475. Yan, H.Y. 1998. Auditory role of the suprabranchial chamber in gourami fish. J. Comp. Physiol. A 183: 325 –333. Zardoya, R. and A. Meyer. 1996. Evolutionary relationships of the coelacanth, lungfishes, and tetrapods based on the 28S ribosomal RNA gene. Proc. Nat. Acad. Sci. USA 93: 5449–5454. Zardoya, R. and A. Meyer. 1997. Molecular phylogenetic information on the identity of the closest living relative(s) of land vertebrates. Naturwiss. 84: 389–397. Zhu, M. and H.-P. Schultze. 2001. Interrelationships of basal osteichthyans. In: Major Events in Early Vertebrate Evolution. Palaeontology, phylogeny, genetics and development (P.E. Ahlberg ed.). Taylor & Francis, New York, pp. 289–314. Zhu, M. and X. Yu. 2002. A primitive fish close to the common ancestor of tetrapods and lungfish. Nature 418: 767–770. Zoology 106 (2003) 3
The ear region of Latimeria chalumnae : functional and evolutionary implications Peter Bernstein Department of Systematic Zoology, Zoological Institute, University of Tübingen, Germany Received November 13, 2002 · Revised version received July 13, 2003 · Accepted September 8, 2003
Summary The anatomy of Latimeria chalumnae has figured prominently in discussions about tetrapod origins. While the gross anatomy of Latimeria is well documented, relatively little is known about its otic anatomy and ontogeny. To examine the inner ear and the otoccipital part of the cranium, a serial-sectioned juvenile coelacanth was studied in detail and a three-dimensional reconstruction was made. The ear of Latimeria shows a derived condition compared to other basal sarcopterygians in having a connection between left and right labyrinths. This canalis communicans is perilymphatic in nature and originates at the transition point of the saccule and the lagena deep in the inner ear, where a peculiar sense end organ can be found. In most gnathostomes the inner ears are clearly separated from each other. A connection occurs in some fishes, e.g. within the Ostariophysi. In the sarcopterygian lineage no connections between the inner ears are known except in the Actinistia. Some fossil actinistians show a posteriorly directed duct lying between the foramen magnum and the notochordal canal, similar to the condition in the ear of Latimeria, so this derived character complex probably developed early in actinistian history. Because some features of the inner ear of Latimeria have been described as having tetrapod affinities, the problem of hearing and the anatomy of the otical complex in the living coelacanth has been closely connected to the question of early tetrapod evolution. It was assumed in the past that the structure found in Latimeria could exemplify a transitional stage in otic evolution between the fishlike sarcopterygians and the first tetrapods in a functional or even phylogenetic way. Here the possibility is considered that the canalis communicans does not possess any auditory function but rather is involved in sensing pressure changes during movements involving the intracranial joint. Earlier hypotheses of a putative tympanic ear are refuted. Key words: Latimeria, inner ear, canalis communicans, Actinistia, Sarcopterygii, early tetrapods
Introduction Since the discovery of the living coelacanth in 1938, many aspects of its morphology have been intensively studied, so that anatomically Latimeria chalumnae is one of the best known fishes in the world today. Nevertheless, some morphological features are still poorly documented, especially those regarding very sensitive soft tissue complexes like the sense organs. Especially the investigation of the inner ear meets with particular difficulties, since its main part consists of very thin membranous canals and compartments, easily de-
stroyed or damaged by pressure changes when catching and pulling the animal from the deeper waters where Latimeria lives. Therefore, so far regrettably little research on the ear of Latimeria has been done. While the innervation pattern of the cranial nerves is known in great detail (Bemis and Northcutt, 1991; Northcutt and Bemis, 1993), only Fritzsch (1987a, b, 1992) examined the innervation of the end organs in the inner ear of Latimeria, reporting a basilar papilla-like structure at the transition point of the saccule and the lagena (Fritzsch, 1987a). These findings have been of systematic interest, since the presence of this putative
Corresponding author: Peter Bernstein, Department of Systematic Zoology, Zoological Institute, University of Tübingen, Auf der Morgenstelle 28, D-72076 Tübingen, Germany; phone: ++49-7071-29-74850, fax: ++49-7071-29-4634, e-mail: [email protected] 0944-2006/03/106/03-233 $ 15.00/0
P. Bernstein
basilar papilla has been used as an important character for defining a sister group relationship between the coelacanth fishes (Actinistia) and the Tetrapoda (Schultze, 1987; Marshall and Schultze, 1992; Zhu and Schultze, 2001), whereas now most scientists agree on a closer relationship between the Tetrapoda and the lungfishes (Dipnoi), using both morphological and molecular data (e.g. Rosen et al., 1981; Meyer and Wilson, 1990, 1991; Cloutier and Ahlberg, 1996; Zardoya and Meyer, 1996, 1997) as shown in Fig. 1. The inner ear of Sarcopterygii
The gross morphology of the inner ear of gnathostomes is quite uniform, in particular the upper part with the three semicircular canals and the utricle. Each of the three semicircular canals houses an ampulla with a sensory crista and is usually connected to the utriculus, which is more or less separated from the sacculus. The lower part with the sacculus and the lagena shows great variability within the Sarcopterygii, especially within the tetrapods. The reason for this complexity is mainly to be found in the evolution of a middle ear and the transformation of certain skeletal elements into ear ossicles, and, as a consequence thereof, the formation of special end organs in the inner ear along with a more elaborate perilymphatic system, functionally coupled with the new middle ear complex. In fishes no middle ear complex has ever evolved. The Weberian apparatus of the Ostariophysi in which modified vertebrae act as a kind of “ear ossicles” between the gas-filled swimbladder and the inner ear may be considered functionally analogous with the middle ear complex. But even without any such structures fishes
Fig. 1. Cladogram of the Sarcopterygii, showing recent crown groups and selected fossil forms (marked with an †). Based on data from Janvier (1996), Cloutier and Ahlberg (1996), Zhu and Schultze (2001) and Zhu and Yu (2002).
234
are not deaf but hear via the macula organs, which act as detectors for linear acceleration and are able primarily to detect passive movements caused in the near field of a sound source (Fritzsch, 2000). In many fishes or fish groups the macula organs do not equally participate in the task of hearing. In the ear of herrings, for example, the macula utriculi is the most important auditory receptor (Popper and Fay, 1993, 1997). Chondrichthyes seem to use the macula neglecta in conjunction with the macula sacculi for hearing (Corwin, 1981). Tetrapods, however, were confronted with a new set of problems when entering the terrestrial environment due to the different physical conditions in an air environment. They developed particular membranous end organs like the basilar papilla or amphibian papilla which in combination with an elaborate system of perilymphatic canals are responsible for the acute sense of hearing in many tetrapods, especially mammals. The middle ear acts as an impedance transducer, overcoming the problems of transferring airborne sound waves to the fluid medium of the inner ear, thus significantly increasing the sensibility of the ear. The evolution of tetrapod hearing
During the past 20 years our knowledge on the evolution of hearing in tetrapods has increased tremendously. Particular attention has been given to the evolutionary transformation of the otical region during the transition from an aquatic to a terrestrial lifestyle. The main interest was directed towards the anatomy of the ear in early tetrapods or members of the tetrapod stem group, since these are transitional forms leading from the fishlike ancestors to the fully terrestrial tetrapods. Many new fossils as well as new interpretations of old findings have revolutionized our knowledge of the vertebrate community at the end of the Devonian and the Early Carboniferous. Some of the fossils are so well preserved (e.g. Acanthostega, Clack, 1989, 1994, 1998) that the anatomy of the ear and related structures can be studied in great detail. So the stapes and its orientation in some very early tetrapods is now well known (Clack, 1983, 1992, 2002). Nevertheless, much remains unknown, because in fossils the soft parts are usually not preserved. This applies not only to the inner ear with its delicate membranous structures, but also to features of the middle ear region such as muscles, nerves, vessels and cartilage. Therefore, a closer look at the more recent tetrapods might be helpful in understanding the evolution of characters in the tetrapod otic region. Unfortunately, this approach has its own limitations. Most living tetrapods have very specialized ears and are poor models for early otic evolution. So, Clack (1998) is surely right in saying that “returning to the earliest tetrapods, I am left with the Zoology 106 (2003) 3
The ear region of Latimeria
overwhelming impression that we must look to fishes such as basal actinopterygians, lungfishes and coelacanths for the nearest analogue of their hearing abilities and inner ear development.” Looking at the living coelacanth, this is all the more important, since some kind of functional tympanic ear was proposed in this taxon (Fritzsch, 1992). The investigation of the auditory function in fishes usually requires electrophysiological, behavioral or even more advanced methods like the auditory brainstem response technique (Kenyon et al., 1998). Because of the elusive lifestyle and the protected status of Latimeria, it seems very unlikely that the coelacanth could be examined in such a way in the foreseeable future. For this reason, we have to check whether the anatomy may provide further insight into the problems mentioned above.
otic capsule is not completely enclosed by cartilage and often lacks a medial wall to the cranial cavity, the inner ear of Latimeria is well encapsulated. The anterior border is situated at the origin of the anteriorly directed processes of the prootic, these being in close contact to the connecting processes of the basisphenoid as part of the intracranial joint. The braincase of Latimeria is divided into an anterior ethmosphenoidal and a posterior otoccipital part. These parts are connected via the intracranial joint, a feature once common in early vertebrates but now restricted to the living coelacanth. All characters involved in hearing are situated in the otoccipital part, which houses the otic capsule and most of the brain.
Materials and methods A juvenile specimen of Latimeria chalumnae Smith, 1939 (CCC-No. 162 K, total length = 310 mm) served as the main object of examination. The specimen had been originally deep-frozen and then preserved in formalin. The state of preservation was good but not perfect. After embedding in celloidin, serial transverse sections 40 µm thick were cut. These were stained with Azan according to Heidenhain. Besides serving as a basis for conventional light microscopic investigation, the serial sections were used to make a three-dimensional reconstruction of the otoccipital part of the cranium using a modified wax plate technique (Born, 1883). Instead of wax, polystyrene was utilized, which is much easier to use. An adult specimen of Latimeria chalumnae Smith, 1939 (CCC-No. 161, SZ 10064), preserved in formalin, was used for estimating the morphometric differences between the juvenile and adult conditions. Both specimens are from the collection of the Department of Systematic Zoology at the University of Tübingen. The specimen is one of 26 embryos that were still inside the mother’s belly at the time of catch, ranging in length from 308 to 358 mm. As far as could be estimated, they were almost ready to be born. The overall proportions are very similar to the adult ones, though a slight lengthening of the skull and the body occurs during later ontogeny.
Results The otic capsule
The otic capsule of Latimeria lies entirely in the otoccipital part of the cranium (Fig. 2 –3). While in many basal actinopterygians and especially in lungfishes the Zoology 106 (2003) 3
Fig. 2. Three-dimensional model of the otoccipital part of a juvenile Latimeria. Parts of the tandem articulation are also shown. Left lateral view. Membrane bones are omitted. The sectional planes of some of the following illustrations are indicated by the respective figure numbers. The region of the thin “tympanumlike” skin is highlighted with a red circle. ah, anterior head of hyomandibula; cartM, Meckelian cartilage; crp, crista parotica; hyom, hyomandibula; inhy, interhyal; opcar, opercular cartilage; ph, posterior head of hyomandibula; pq, palatoquadrate; pro, connecting processes of prootic; qd, quadrate; symp, symplectic; styl, stylohyal.
235
P. Bernstein
Fig. 3. Skull of an adult Latimeria chalumnae. Left lateral view. Redrawn from Millot et al. (1978). Abbreviations see Fig. 2.
Fig. 5. Transverse section through the skull of a juvenile Latimeria chalumnae. On the right hand side, the foremost part of the spiracular pouch lies between the connecting processes of the prootic and the expansive palatoquadrate. The thin skin between the spiracular pouch and the outer medium is marked with a red circle. bm, basicranial muscle; br, brain; bsph, basisphenoid; eptg, entopterygoid; icav, intracranial cavity; itmp, intertemporal; mcav, mouth cavity; mptg, metapterygoid; nc, notochord; pq, palatoquadrate; pro, connecting processes of prootic; spc, spiracular chamber.
Fig. 4. Transverse section through the skull of a juvenile Latimeria chalumnae. The otic capsule is completely enclosed by cartilage except for the medial foramen of the canalis communicans, which originates at the end organ. bm, basicranial muscle; br, brain; cc, canalis communicans; crc, crus commune; eo, end organ; for, foramen in medial wall of otic capsule; icav, intracranial cavity; mcav, mouth cavity; nc, notochord; ophy, opercular process of hyomandibula; stmp, supratemporal.
The dorsolateral wall of the otic capsule is well developed, particularly at the hyomandibular insertion point. Here, a prominent lateral crest (crista parotica) or otic shelf strengthens the articulation area. Medially, the inner ear is separated from the notochord and the cranial cavity by a partially ossified medial wall, which is perforated by foramina for nerves and other elements 236
such as the endolymphatic duct. Ossification of the medial wall is to be seen close to the notochord only, obviously reinforcing the otic capsule against forces active during movements within the intracranial joint. The same applies to the floor of the capsule at the inserting region of the basicranial muscle, where the skeletal elements have to withstand strong forces during contraction of this muscle (Fig. 4). The spiracular pouch and the hyomandibula
Latimeria has a spacious spiracular pouch but no spiracular opening. At the level of the intracranial joint, where the upper portion of the palatoquadrate lies, the anterior end of the spiracular pouch is dorsally separated from the outside by a thin membrane-like skin (Fig. 2, 5). It extends backwards and is in broad contact with the mouth cavity at the level where the prootic processes connect to the otoccipital roof. The posterior margin extends to the space above the lateral crest where it reaches the anterior articulation of the hyomandibula. Zoology 106 (2003) 3
The ear region of Latimeria
The hyomandibula has four articulations with other skeletal elements. Two articulations connect the hyomandibula to the braincase, a posteriorly directed process articulates to the opercular, and a ventral process contacts the interhyal. The anterior, more ventral articulation to the otic capsule is associated with strong ligamentous connective tissue, whereas the posterior, more dorsal insertion shows weaker ligaments. The inner ear
Fig. 6. The right inner ear of an adult Latimeria chalumnae in medial view. From Millot and Anthony (1965), modified. Note the dissected canalis communicans. The position of the end organ is marked in red. cc, canalis communicans; dend, endolymphatic duct; eo, end organ; lag, lagenar recess; sac, sacculus. VIII, cranial nerve 8.
The inner ear of Latimeria was described in detail by Millot and Anthony (1965). The upper part shows the usual gnathostome pattern with a moderate-sized utricle and three perpendicularly oriented semicircular canals (Fig. 6). In the lower part the sacculus is very prominent. It houses a very large otolith on the medial side and is connected to a posteriorly directed and separated lagenar recess. At the transition point of the saccule and the lagenar recess an orifice is situated, marking the origin of a bilateral connection between the inner ears, known as canalis communicans. Its exact origin is situated where a peculiar membrane-like structure can be found deeper in the inner ear (Fig. 4, 7– 8). From there the canal passes medially through a foramen in the medial wall of the otic capsule and extends backwards between the notochord and the otic capsule. For most of its course the canalis communicans is completely enclosed by cartilage (Fig. 9).
Fig. 7. The canalis communicans and the end organ in the inner ear of Latimeria chalumnae. This figure shows an enlarged view of a similar plane as Fig. 4 in histological display. cc, canalis communicans; eo, end organ; for, foramen in medial wall of otic capsule; nc, notochord; sac, sacculus.
Fig. 8. A further enlarged view shows the peculiar end organ in the inner ear of Latimeria chalumnae lying in the endolymphatic space. Above, the origin of the canalis communicans is situated. cc, canalis communicans; eo, end organ; es, endolymphatic space; ps, perilymphatic space.
Zoology 106 (2003) 3
237
P. Bernstein
canals disappear not far from that place, both parts of the canalis communicans pass further backwards, eventually merging in the occipital region between the foramen magnum and the notochord (Fig. 9D). Where the canalis communicans enters the saccular/lagenar region, an apparent end organ with a membrane spans the orifice between the endo-/perilymphatic spaces (Fig. 7– 8). Clearly, this is the structure regarded by Fritzsch (1987a) as a basilar papilla. It seems indeed to be an innervated end organ, consisting of a membrane and possibly an overlying group of cells (Fig. 8). Its size is about 1 mm and it seems to span the entire distance between the canalis communicans and the saccular/lagenar orifice. It is apparently innervated by fibers from the lagenar branch of the statoacustic nerve VIII. There are two completely separated spaces, one endolymphatic and one perilymphatic, with the end organ lying in the endolymphatic portion (Fig. 8). The perilymphatic space is enclosed by a duct (shown collapsed in Fig. 7) which traverses the medial wall of the otic capsule and passes further backwards without forming an expanded pouch as depicted by Fritzsch (1992). Instead, the foramen in the medial wall is comparatively narrow with the duct passing immediately posteriorly after traversing this foramen (Fig. 4).
Discussion A tympanic ear in Latimeria?
Fig. 9. Four consecutive serial sections show the passage of the canalis communicans in posterior direction. The connection of the lower and upper canals is marked by an arrow. cc, canalis communicans; icav, intracranial cavity; nc, notochord; ucc, upper part of canalis communicans.
Not far behind the foramen for the canalis communicans another set of canals originates at the floor of the cranial cavity, dorsally of the canalis communicans (Fig. 9A). These canals become enclosed by cartilage on their passage backwards, so that in cross-sections a double set of canals can be seen for quite a distance (Fig. 9B–D). Near the posterior end of the otic capsule the upper canals and the canalis communicans are connected by a very tiny duct on each side (Fig. 9C). While the upper 238
The identification of a papilla-like receptor end organ in the inner ear of Latimeria and the finding of a thin membrane-like skin between the spiracular pouch and the outer medium led Fritzsch (1992) to hypothesize a tympanic function of the auditory system in the coelacanth. The spiracular pouch would in this case act as a pressure-detection system with the thin skin functioning as a vibrating membrane much as the tympanum in tetrapods. Furthermore, Fritzsch (1992) proposed several ways how pressure changes could be transmitted to the receptor end organ he interpreted as a basilar papilla. These include possible pressure differences between the spiracular pouch and the intracranial cavity, pressure differences between the spiracular pouch and the belly transmitted by the canalis communicans, participation of the lower jaw and the hyomandibula in inducing pressure differences between the inner ear and the brain cavity and movements within the intracranial joint. To evaluate these hypotheses one has to look at the function of the intracranial joint and the hyomandibula, the anatomy of the inner ear and the canalis communicans and the relationships among those structures. In order to act as a resonance chamber, the spiracular Zoology 106 (2003) 3
The ear region of Latimeria
pouch of Latimeria would have to be filled with gas. The mere existence of a gas-filled chamber like the swimbladder somewhere in the body improves the hearing abilities in fish (Blaxter, 1981), and the closer this gas-filled chamber is to the ear the better. In anabantoid fishes (Schneider, 1941; Yan, 1998) the improvement of hearing capabilities by air-filled chambers near the otic capsules is well known. Fishes in other taxonomic groups have developed other means of auditory enhancement. These include the clupeids, with otic bullae connected to the swimbladder by small tubes, and the Ostariophysi (e.g. carp, catfishes), with a series of bony elements between the swimbladder and the ear, known as the Weberian apparatus. These bony elements act as “ear ossicles”, transmitting sound-generated vibrations from the swimbladder to the inner ear. The superficial resemblance of the inner ear connection in the Ostariophysi and Latimeria was already noticed by Millot and Anthony (1965), but they also pointed out the lack of any structures transmitting pressure waves to the ear of the coelacanth and the impossibility of the coelacanth’s fat-filled swimbladder to act as a resonance chamber. Moreover, the swimbladder is not situated in the vicinity of the ear. There is no obvious sound-transmitting apparatus from the abdominal region to the canalis communicans. Fritzsch (1992) pointed out the discovery of some fossil actinistians with an apparently ossified swimbladder and its resemblance to the ossified swimbladders of some recent ground-dwelling catfishes (Chranilov, 1929; Fiedler, 1991). Indeed, these catfishes may use their swimbladder as a resonance chamber for hearing, but one has to keep in mind that these fishes are ostariophysines and possess the Weberian apparatus, which is completely lacking in Latimeria. There is experimental evidence that these catfishes with an ossified swimbladder-wall even have decreased hearing ability (Ladich, 1999), so the discovery of ossified swimbladders in fossil actinistians probably does not indicate a pathway for pressure waves from the swimbladder to the inner ear, rather the opposite. The ossification of the swimbladder probably is an adaptation in ground-dwelling or deep-sea fishes to better withstand the high water pressure. Since gas is easily compressed, it would be hard for any fish living at greater depths to maintain the gas filling in its swimbladder. The same might be true for Latimeria, where the swimbladder is fat-filled. The fatty substance gives buoyancy and is not easily compressed in depths of 150 –700 m, where Latimeria usually lives (Fricke and Plante, 1988). Furthermore, the existence of an ossified swimbladder or lung in the Actinistia seems to be restricted to Mesozoic forms like Undina or Macropoma (Janvier, 1996). Since the Devonian genus Diplocercides already shows a canalis communicans (Fig. 10) but no ossified swimZoology 106 (2003) 3
bladder (Jarvik, 1980), a functional correlation between these two characters does not seem very likely. Another problem regarding a possible gas-filled spiracular pouch is the great depth Latimeria lives in. As noted above, it is not easy to maintain a gas-filled cavity against the high water pressure in deep water. And how would this cavity be filled with gas? Only two explanations would be reasonable: either Latimeria has to take up air at the surface or there has to be a gas-secreting gland in the spiracular pouch. While there is no evidence for the latter possibility, the first is not supported by observation either. The depth Latimeria lives in (ca. 200 m and deeper) does not support such a notion. The role of the hyomandibula
The use of a gas-filled spiracular chamber near the ear capsule and the hyomandibula as a receptor of far-field sound in fishes has been discussed by van Bergeijk (1966). The proposed close functional connection between these elements is in line with the well established homologies between the hyomandibula and the stapes on the one hand and between the spiracular pouch and the tetrapod middle ear cavity on the other hand. In the traditional view of otic evolution during the transition from water to land, the position of the hyomandibula close to the ear capsule and its vicinity to the spiracular pouch made for an easy transformation of these elements into parts of an air-adapted tetrapod ear (Lombard and Bolt, 1979). Although this view has been challenged by the discovery of bulky stapes in the earliest tetrapods, which were seemingly quite inappropriate for aerial hearing and suitable for sensing ground vibra-
Fig. 10. Reconstructed brain case of the Devonian actinistian Diplocercides, occipital view. The canalis communicans is situated in the same position as in Latimeria, between the foramen magnum and the notochordal canal. Redrawn from Jarvik (1980). cc, canalis communicans; fm, foramen magnum; nc, notochord.
239
P. Bernstein
tions only (Clack, 1992), the possibility remains that the condition of these structures in Latimeria could present a transitory stage in a functional, or even phylogenetic context, if the conditions in Latimeria are interpreted as ancestral as proposed by Fritzsch (1992). According to Eaton (1939), in the early sarcopterygians the hyomandibula originally had five articulations with other parts of the skull: two heads inserting at the otic capsule, an opercular process, an articulation with the quadrate and another articulation with the ceratohyal. If we accept this as the sarcopterygian groundplan, Latimeria has four of these articulations and has lost the one to the quadrate, since the quadrate-Meckelian cartilage joint is separated from the symplectic-Meckelian cartilage joint due to the peculiar tandem articulation of the coelacanth’s jaws. While this constellation is surely derived, the hyomandibula of Latimeria is part of a complex skull kinesis and certainly has an important function in integrating the movements of the various skull elements, probably just as in the extinct “osteolepiforms” (Thomson, 1967). Only after the hyomandibula lost this function, it could be used in a different functional context and got involved in hearing (Thomson, 1966b). Although there are quite a few incompatible theories on the skull kinetics of Latimeria (e.g. Thomson, 1966a; Robineau and Anthony, 1973; Alexander, 1973; Adamicka and Ahnelt, 1976; Lauder, 1980), it seems to be clear that the hyomandibula of Latimeria is in no way “unfunctional”, but participates in the complicated sequence of movements during mouth opening and closing, prey capture and breathing. The strong attachment of the hyomandibula to the otic capsule, allowing only limited mobility, was already noted by Alexander (1973). The existence of a robust ligamentous connection of the anterior hyomandibular head articulating to the otic capsule seems to support this view. A hingelike movement of the hyomandibula against the braincase is all that seems to be possible. The lateral wall of the otic capsule where the medial heads of the hyomandibula insert is especially thick, apparently to strengthen this critical and strained location. The perilymphatic space connected to the end organ in the inner ear is on the opposite side near the medial wall of the otic capsule. A functional connection between the perilymphatic canalis communicans and the hyomandibula seems very unlikely. A similar situation can be found regarding a possible connection of the spiracular pouch and the hyomandibula. The latter is far away from the end organ or the canalis communicans and only in peripheral contact with the posteriormost offshoot of the spiracular pouch. The obvious conclusion is that there is no functional connection between the hyomandibular apparatus and the inner ear end organs or the canalis communicans. 240
The canalis communicans and the “basilar papilla”
The anatomy and the location of the mentioned elements do not support a functional relationship (at least not an auditory one) between those elements and the canalis communicans and the peculiar end organ in the inner ear. There has to be another function of this inner ear connection, which has obviously been present since the early actinistian evolution, as the skeletal canalis communicans seemed to be already present in the Devonian actinistian Diplocercides (Jarvik, 1980). As far as can be guessed from the description of the reconstructed brain case, Diplocercides had a canalis communicans very similar to the coelacanth’s one (Fig. 10). Unfortunately, this canal is not found in other fossil actinistians. That does not necessarily mean that there was none, because in most fossil actinistians the region surrounding the canalis communicans was obviously cartilaginous and therefore not preserved (Forey, 1998). The canalis communicans originates at the observed end organ deep in the inner ear and reaches the foramen in the medial wall of the otic capsule via a duct and not via a perilymphatic sac as proposed by Fritzsch (1992). There is no connection between this duct and the undifferentiated perilymphatic space surrounding the sacculus. The main connection to the intracranial space is not at the place of the perilymphatic foramen in the medial wall but via tiny ducts leading to the upper set of canals paralleling the canalis communicans proper (Fig. 9). Those ducts are connected to the intracranial space and are surely able to transmit any pressure changes inside the intracranial cavity to the inner ear. The possibility of transmission of pressure changes via this pathway during movements within the intracranial joint was recognized by Fritzsch (1992), but in this case an auditory function of the whole complex must be questioned. For any auditory specialization (pressure detection and transduction of displacement) there has to be a low-density pathway between the hair cells of the end organ and a displacement transducer via a periotic fluid system as found in the tetrapod ear and the ears of some specialized teleosts (Lombard and Hetherington, 1993), which could not be found in the living coelacanth. There is no evidence for any participation of the intracranial joint in hearing. If the end organ is sensitive to pressure changes within the intracranial cavity, it might be part of controlling the complex pattern of movements involved in the jaw mechanics of Latimeria. In this case it would be rather a specialized organ possibly unique to the actinistian lineage and not homologous to the basilar papilla in tetrapods. Taking into account the points mentioned above, we have to conclude that the ear of Latimeria is hardly suitable as a functional transitional form between the ear of early fishlike sarcopterygians and the modern tetrapod ear. Zoology 106 (2003) 3
The ear region of Latimeria
These results seem to show once more the complexity of the character mosaic of Latimeria that every systematist is confronted with. While some characters of the coelacanth apparently are plesiomorphic such as the structure of the brain (Millot and Anthony, 1965), the secondary capillary system in the gills (Vogel et al., 1998) or the myoseptal architecture of the trunk muscles (Gemballa and Ebmeyer, 2003), others are clearly autapomorphic to Latimeria or actinistians as a group, like the rostral organ (Bemis and Hetherington, 1982) or the unique tandem articulation of the jaws (Adamicka and Ahnelt, 1976). Concerning the inner ear of Latimeria, the little research previously done indicates a plesiomorphic condition of the hair cell pattern in the macula lagenae, without any affinities to the tetrapod state (Platt, 1994). The end organ at the origin of the canalis communicans may not be similar to the basilar papillae found in basal tetrapods such as salamanders or plesiomorphic anurans. In the ears of these forms, the basilar papillae as well as the amphibian papillae are easily identified by their location and appearance, with the hair cells conspicuously distinguishable. The end organ of Latimeria does not easily allow such a homologization and is morphologically different from the papillae of lissamphibians. Fritzsch (1987, 1992) is right in pointing out the similarity of the end organ’s innervation by a twig of the lagenar branch of the VIIIth nerve to the innervation pattern of the basilar papilla in tetrapods. But does that necessarily mean that both structures are homologous? We have evidence that the sensory epithelia in the vertebrate ears developed by the splitting of a single sensory precursor (Fritzsch et al., 2002). If so, the end organ in the ear of Latimeria could be the part representing the homologon to the basilar papilla in tetrapods, but because there is also evidence for the independent origin of the lagena (Fritzsch, 1992; Fritzsch et al., 2002) this could likewise point to an independent origin of the end organ coupled to the lagena. This would be the basilar papilla in tetrapods but could be another end organ in other vertebrate groups, e.g. the coelacanth. The basilar papilla in tetrapods is coupled to an auditory function; using the term “basilar papilla” for an end organ found in the inner ear of a vertebrate could easily tempt one to assume an auditory function similar to the ear of tetrapods. As long as we have no evidence for such a function of the end organ in the coelacanth’s ear, I would recommend to avoid the term “basilar papilla”. In a phylogenetic sense the results support the uniqueness of the coelacanth’s cranial functional morphology, even more so if the end organ and the canalis communicans are involved in controlling the mechanics of the intracranial joint rather than in hearing. The condition Zoology 106 (2003) 3
of the Devonian coelacanth Diplocercides probably shows us the early evolution of this character. Therefore, the end organ in the inner ear of Latimeria may not be represented as synapomorphic with the tetrapods. For this reason I prefer the phylogeny of living sarcopterygians as shown in Fig. 1, showing the dipnoi as closely related to the tetrapods, with the actinistians as the sister group of the clade formed by the two other groups.
Acknowledgements I would like to thank Prof. Dr. W. Maier for providing access to the material of Latimeria and the permission to use the facilities at the Lehrstuhl für Spezielle Zoologie at the University of Tübingen. The specimens of Latimeria were generously provided by Prof. Dr. H. Fricke. Many helpful discussions with S. Gemballa, M. Hoffmann and M. Sánchez-Villagra provided invaluable advice. M. Sánchez-Villagra read an earlier draft of this paper and gave many useful comments. I also want to thank an anonymous reviewer and Bernd Fritzsch for many helpful remarks. K. Hagen was very helpful in translating parts of the work of Millot and Anthony (1965) from French into German. M. Meinert, T.-T. Fußnegger and G. Schmid provided technical assistance. This study was supported by a grant from the Landesgraduiertenförderung Baden-Württemberg.
References Adamicka, P. and H. Ahnelt. 1976. Beiträge zur funktionellen Analyse und zur Morphologie des Kopfes von Latimeria chalumnae Smith. Ann. Naturhist. Mus. Wien 80: 251–271. Alexander, R.M. 1973. Jaw mechanisms of the Coelacanth Latimeria. Copeia 1: 156 –158. Bemis, W.E. and T.E. Hetherington. 1982. The rostral organ of Latimeria chalumnae: morphological evidence of an electroreceptive function. Copeia 2: 467– 471. Bemis, W.E. and R.G. Northcutt. 1991. Innervation of the basicranial muscle of Latimeria chalumnae. Environ. Biol. Fish. 32: 147–158. Bergeijk, W.A. van. 1966. Evolution of the sense of hearing. Amer. Zool. 6: 371–377. Blaxter, J.H. S. 1981. The swimbladder and hearing. In: Hearing and Sound Communication in Fishes (W. N. Tavolga, A.N. Popper and R.R. Fay, eds.). Springer, New York, pp. 61–71. Born, G. 1883. Die Plattenmodellirmethode. Arch. Mikros. Anat. 22: 584–599. Clack, J.A. 1983. The stapes of the Coal Measures embolomere Pholiderpeton scutigerum Huxley (Amphibia: Anthracosauria) and otic evolution in early tetrapods. Zool. J. Linn. Soc. 79: 121–148. Clack, J.A. 1989. Discovery of the earliest-known tetrapod stapes. Nature 342: 425–427. Clack, J.A. 1992. The stapes of Acanthostega gunnari and the role of the stapes in early tetrapods. In: The Evolutionary Biol-
241
P. Bernstein
ogy of Hearing (D. B.Webster, R. R. Fay and A. N. Popper, eds.). Springer, New York, pp. 405 – 420. Clack, J.A. 1994. Earliest known tetrapod braincase and the evolution of the stapes and fenestra ovalis. Nature 369: 392–394. Clack, J.A. 1998. The neurocranium of Acanthostega gunnari Jarvik and the evolution of the otic region in tetrapods. Zool. J. Linn. Soc. 122: 61– 97. Clack, J.A. 2002. Gaining Ground: The Origin and Evolution of Tetrapods. Indiana University Press, Bloomington. Cloutier, R. and P.E. Ahlberg. 1996. Morphology, characters, and the interrelationships of basal sarcopterygians. In: Interrelationships of Fishes (M.L.J. Stiassny, L.R. Parenti and G.D. Johnson, eds.). Academic Press, San Diego, pp. 445 –479. Corwin, J.T. 1981. Audition in Elasmobranchs. In: Hearing and Sound Communication in Fishes (W.N. Tavolga, A.N. Popper and R.R. Fay, eds.). Springer, New York, Berlin, Heidelberg, pp. 81–105. Eaton, T.H. 1939. The crossopterygian hyomandibular and the tetrapod stapes. J. Wash. Acad. Sci. 29: 109 –117 Fiedler, K. 1991. Lehrbuch der speziellen Zoologie. Band 2: Wirbeltiere, Teil 2: Fische. Fischer, Stuttgart. Forey, P.L. 1998. History of the Coelacanth Fishes. Chapman & Hall, London. Fricke, H., O. Reinicke, H. Hofer and W. Nachtigall. 1987. Locomotion of the coelacanth Latimeria chalumnae in its natural environment. Nature 329: 331– 333. Fricke, H. and R. Plante. 1988. Habitat requirements of the living coelacanth Latimeria chalumnae at Grande Comore, Indian Ocean. Naturwiss. 75: 149 –151. Fritzsch, B. 1987a. Inner ear of the coelacanth fish Latimeria has tetrapod affinities. Nature 327: 153 –154. Fritzsch, B. 1987b. Die Papilla basilaris von Latimeria chalumnae und die Evolution von amphibientypischen auditiven Sinnesepithelien. Verh. Dtsch. Zool. Ges. 80: 145 –146. Fritzsch, B. 1992. The water-to-land transition: evolution of the tetrapod basilar papilla, middle ear, and auditory nuclei. In: The Evolutionary Biology of Hearing (D.B. Webster, R.R. Fay and A.N. Popper, eds.). Springer, New York, pp. 351–375. Fritzsch, B. 2000. Microscopic functional anatomy: Sensory systems: Hearing. In: The Laboratory Fish (G.C. Ostrander, ed.). Academic Press, London, pp. 480 – 487. Fritzsch, B., K.W. Beisel, K. Jones, I. Fariñas, A. Maklad, J. Lee and L.F. Reichardt. 2002. Development and evolution of inner ear sensory epithelia and their innervation. Journal of Neurobiology 53: 143 –156. Gemballa, S. and L. Ebmeyer. 2003. Myoseptal architecture of sarcopterygian fishes and salamanders with special reference to Ambystoma mexicanum. Zoology 106: 29 – 41. Janvier, P. 1996. Early Vertebrates. Clarendon Press, Oxford. Jarvik, E. 1980. Basic Structure and Evolution of Vertebrates. Academic Press, London. Kenyon, T.N., F. Ladich and H.Y. Yan. 1998. A comparative study of hearing ability in fishes: the auditory brainstem response approach. J. Comp. Physiol. A 182: 307– 318. Ladich, F. 1999. Did auditory sensitivity and vocalization evolve independently in otophysan fishes? Brain Behav. Evol. 53: 288 – 304. Lauder, G.V. 1980. The role of the hyoid apparatus in the feeding mechanism of the Coelacanth Latimeria chalumnae. Copeia 1: 1– 9. Lombard, R.E. and J.R. Bolt. 1979. Evolution of the tetrapod ear: an analysis and reinterpretation. Biol. J. Linn. Soc. 11: 19–76. Lombard, R.E. and E.H. Hetherington. 1993. Structural basis of hearing and sound transmission. In: The Vertebrate Skull, Volume 3: Functional and Evolutionary Mechanisms (J. Hanken
242
and B.K. Hall, eds.). University of Chicago Press, Chicago, pp. 241–302. Marshall, C. and H.-P. Schultze. 1992. Relative importance of molecular, neontological, and paleontological data in understanding the biology of the vertebrate invasion of the land. J. Mol. Evol. 35: 93–101. Millot, J. and J. Anthony. 1965. Anatomie de Latimeria chalumnae. II. Système nerveux et organes des sens. C.N.R.S., Paris. Millot, J., J. Anthony and D. Robineau. 1978. Anatomie de Latimeria chalumnae. III. Appareil digestif, appareil respiratoire, appareil uro-génital, glandes endocrines, appareil circulatoire, téguments, écailles, conclusions générales. C.N.R.S., Paris. Meyer, A. and A.C. Wilson. 1990. Origin of tetrapods inferred from their mitochondrial DNA affiliation to lungfish. J. Mol. Evol. 31: 359–364. Meyer, A. and A.C. Wilson. 1991. Coelacanth’s relationships. Nature 353: 219. Northcutt, R.G. and W.E. Bemis. 1993. Cranial nerves of the coelacanth, Latimeria chalumnae (Osteichthyes: Sarcopterygii: Actinistia), and comparisons with other Craniata. Brain Behav. Evol. 42 [Suppl. 1]: 1–74. Platt, C. 1994. Hair cells in the lagenar otolith organ of the coelacanth are unlike those in amphibians. J. Morph. 220: 381. Popper, A.N. and R.R. Fay. 1993. Sound detection and processing by fish: critical review and major research questions. Brain Behav. Evol. 41: 14 –38. Popper, A.N. and R.R. Fay. 1997. Evolution of the ear and hearing: issues and questions. Brain. Behav. Evol. 50: 213–221. Robineau, D. and J. Anthony. 1973. Biomécanique du crâne de Latimeria chalumnae (Poisson crossoptérygien coelacanthidé). C.R. Acad. Sci. Paris Sér. D 276: 1305–1308. Rosen, D.E., P.L. Forey, B.G. Gardiner and C. Patterson. 1981. Lungfishes, tetrapods, paleontology, and plesiomorphy. Bull. Am. Mus. Nat. Hist. 167: 159 –276. Schneider, H. 1941. Die Bedeutung der Atemhöhle der Labyrinthfische für ihr Hörvermögen. Z. vergl. Physiol. 29: 172 –194. Schultze, H.-P. 1987. Dipnoans as Sarcopterygians. J. Morphol. Suppl. 1: 39 –74. Thomson, K.S. 1966a. Intracranial mobility in the coelacanth. Science 153: 999 –1000. Thomson, K.S. 1966b. The evolution of the tetrapod middle ear in the rhipidistian-amphibian transition. Amer. Zool. 6: 379 –397. Thomson, K.S. 1967. Mechanisms of intracranial kinetics in fossil rhipidistian fishes (Crossopterygii) and their relatives. Zool. J. Linn. Soc. 46: 223–253. Vogel, W.O.P., G.M. Hughes and U. Mattheus. 1998. Non-respiratory blood vessels in Latimeria gill filaments. Phil. Trans. R. Soc. Lond. B 353: 465– 475. Yan, H.Y. 1998. Auditory role of the suprabranchial chamber in gourami fish. J. Comp. Physiol. A 183: 325 –333. Zardoya, R. and A. Meyer. 1996. Evolutionary relationships of the coelacanth, lungfishes, and tetrapods based on the 28S ribosomal RNA gene. Proc. Nat. Acad. Sci. USA 93: 5449–5454. Zardoya, R. and A. Meyer. 1997. Molecular phylogenetic information on the identity of the closest living relative(s) of land vertebrates. Naturwiss. 84: 389–397. Zhu, M. and H.-P. Schultze. 2001. Interrelationships of basal osteichthyans. In: Major Events in Early Vertebrate Evolution. Palaeontology, phylogeny, genetics and development (P.E. Ahlberg ed.). Taylor & Francis, New York, pp. 289–314. Zhu, M. and X. Yu. 2002. A primitive fish close to the common ancestor of tetrapods and lungfish. Nature 418: 767–770. Zoology 106 (2003) 3