Course Hero - We put you ahead of the curve!
You have requested the below document.
- Title: SchwenkWakeBJLS93
- Type: Notes
- School: UConn
- Course: EEB 349
- Term: Fall
Journal Biological o the I.innean Socieb (1993), 49: 141-162. With 3 figures f Prey processing in Leurognathus rnarrnoratus and the evolution of form and fmction in desmognathine salamanders (Plethodontidae) KURT SCHWENK Department o Ecology and Evolutionary Biology, University o Connecticut, f f Storrs, CT 06269-3043, U.S.A. AND DAVID B. WAKE Museum o Vertebrate <oology and Department o Integrative Biology, f f University o Calzfornia, Berkeley, CA 94720, U.S.A. f Received 23 Juh 1991, accepted for publication 16 March 1992 Function and biological role of morphological specialization in desmognathine salamanders are analysed in the light of studies of feeding in Lcurognathus marmoratus. Nine morphological features uniquely characterize the Desmognathinae as compared to its sister group, the Plethodontinae, and other salamanders: ( 1 ) heavily ossified and strongly articulated skull and mandible; (2) flat, wedgelike head profile; (3) stalked occipital condyles; (4) modified atlas; (5) modified anterior trunk vertebrae; (6) atlanto-mandibular ligaments; (7) enlarged dorsal spinal muscles; (8) enlarged quadrato-pectoralis muscles; and (9) hind limbs relatively larger than forelimbs. Dorsoventral head mobility is increased at the atlanto-occipital joint by the stalked occipital condyles which simultaneously increase the mechanical advantage of the hypertrophied axial muscles that cross the joint. During head depression the atlanto-mandibular ligaments are placed in tension. Force generated by the quadrato-pectoralis muscles is transmitted directly to the mandible, creating a powerful bite with the jaws in full occlusion. Desmognathines use an efficient static pressure system for subduing and/or killing prey items held in the jaws, not a kinetic-inertial mechanism, as previously suggested. Lcurognathus exhibits a behaviour ( head-tucking ) unique to desmognathines that is consistent with the static-pressure hypothesis. Several desmognathine features ( I , 2, 5, 7, 9) are not explicable as adaptations for feeding; these function as locomotory specializations for burrowing, especially for wedging under rocks within and alongside streams. Desmognathines use head-tucking during such wedging and burrowing movements, thus locomotory specializations act in concert with the feeding specializations. We suggest that origin of the atlanto-mandibular ligaments can be considered a key innovation in that it allowed the secondary invasion of stream habitats by adults of ancestral desmognathines. ADDITIONAL KEY WORDS:-Urodela - morphology - evolution. - Desmognathinae - feeding - locomotion - burrowing CONTENTS Introduction. . . Morphological characters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 143 Correspondence to K. Schwenk. 0024-4066/93/060141+ 22 $08.00/0 141 0 1993 The Linnkan Society of London I42 K. SCHWENK AND D. B. WAKE Heavily ossified and strongly articulated skull and mandible Flat, wedgelike head profile . . . . . . . . Stalked occipital condyles ........ Modified atlas . . . . . . . . . . . Modified anterior trunk vertebrae ...... Atlanto-mandibular ligaments ....... Enlarged dorsal spinal muscles ....... Enlarged quadrato-pectoralis muscles . . . . . . Hind limbs relatively larger than forelimbs .... Review of putative function and biological role .... Prey processing in Lcurognalhus marmoralus . . . . . . Background and methods ........ Prey processing in Lcurognalhus ....... Discussion ............. Desmognathine specialization and feeding .... Desmognathine specialization and burrowing . . . . Evolutionary conclusions . . . . . . . . . Acknowledgements . . . . . . . . . . . References ............. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . , . . ... ... ... ... ... . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 143 143 144 144 144 145 145 145 145 149 149 149 151 153 156 158 159 160 INTRODUCTION The Desmognathinae comprises three genera of unusual plethodontid salamanders. Desmognathus Baird, 1880, Leurognathus Moore, 1899 and Phaeognathus Highton, 1961. The distinctiveness of this subfamily was emphasized historically by its elevation to familial status (e.g. Cope, 1869, 1889; Smith & Taylor, 1948; Soler, 1950), but current consensus includes it within Plethodontidae (e.g. Dunn, 1926; Wake, 1966; Regal, 1966; Lombard & Wake, 1986; Duellman & Trueb, 1986). This disagreement has remained purely taxonomic; no worker has doubted the monophyly of Desmognathinae nor its sister relationship with the plethodontine plethodontids. The principal basis for both the taxonomic disagreement noted above and the strong confidence evinced in desmognathine monophyly is the shared presence in this group of a uniquely derived suite of morphological characters. These include the following: (1) heavily ossified and strongly articulated skull and mandible; (2) streamlined, flat, wedgelike head profile; (3) stalked (pedunculate) occipital condyles; (4) modified atlas; (5) modified anterior trunk vertebrae; (6) atlanto-mandibular ligaments; (7) hypertrophied dorsal spinal muscles; (8) hypertrophied quadrato-pectoralis (gularis) muscles; and (9) hind limbs relatively larger than forelimbs. The morphological manifestations of these characters have been dealt with by a number of workers (see below), most comprehensively by Wake (1966). These authors are remarkably uniform in their accounts. In contrast, the functional significance of the characters remains uncertain and is the source of some contention. In particular, recent functional analyses of salamander feeding have challenged earlier conclusions based on morphological inference (Eaton, 1957; Larsen & Guthrie, 1975; Dockx & de Vree, 1986; Dalrymple, Juterbock & La Valley, 1985; Larsen & Beneski, 1988). Furthermore, most workers have dealt with a limited subset of salient characters, while other have assumed a piori that all are functionally related and, therefore, shaped by a single cause. Lastly, there has been general confusion regarding the function of a feature versus its biological role (Bock & von Wahlert, 1965). Consequently, some disagreements in the literature reflect differing conclusions about biological role (e.g. feeding vs SALAMANDER FORM AND FUNCTION 143 burrowing) based on essentially similar functional conclusions (e.g. increased cranial dorsoflexion). In this paper we review the morphology of the characters in question and the history of ideas regarding their functional and biological interpretation. We present new functional data on feeding in Leurognathus marmoratus that bear directly on these issues, and we analyse these data in light of previously published hypotheses. MORPHOLOGICAL CHARACTERS Heavily ossified and strongly articulated skull and mandible Soler (I950), Wake (I966),Hinderstein (1971). The skull and particularly the mandible is heavier and more solid than in plethodontine species. In general, skull bones are thicker and more closely articulated. T h e suspensoria are ossified much more extensively in desmognathines than in plethodontines and bony articulations are broader and firmer. For example, left and right parietal bones interlock by means of numerous finger-like projections along their articulating surfaces. The parietal and squamosal articulate tightly by means of a tongueand-groove joint. These types of articulations are unusual among salamanders. The premaxillae are particularly massive; their frontal processes are fused, leaving only a small internasal fontanelle which is lost entirely in some old individuals. The processes are broad and only slightly narrower than the premaxillary tooth row, and the palatal shelves are broader and better developed than in any plethodontine. The parasphenoids are uniquely hypertrophied in desmognathines. They are broad posteriorly, with large posterolateral processes that extend to a n articulate with the quadrates. This morphology is diagnostic of the Desmognathinae. The prefrontals are absent as discrete elements, allowing for a broad and solid articulation between the maxilla and frontal. The frontal has a large, bony process that extends into the antorbital region and apparently strengthens the skull in the region between the vomer and the frontal-maxilla. Cope (1869) found the ossified antorbital to be so striking that he used it as a diagnostic character for his Desmognathidae. The dentaries are large, massive, and broad ventrally. The Meckelian grooves are nearly closed and the dentary encases the prearticulars. Flat, wedgelike head proJile Soler (1950), Wake (1966). In lateral view the head ofdesmognathines is flat and wedgelike. The dorsal surface of the skull is flat, smooth and devoid of muscles, which have shifted their origins lateroventrally relative to plethodontines. Some parts of the adductor musculature (e.g. the adductor mandibulae anterior) lie in a deep groove or trough. T h e dorsolateral margins of the skull are ridgelike. Vomerine vaulting is absent or reduced. Slalked occipital condyles Dunn ( I 9 2 6 ) , Noble (1927, 1931), Soler (1950), Wake (1966), Hinderstein (1971). The occipital condyles of all non-desmognathine salamanders are sessile , i.e. 144 K. SCHWENK AND D. B. WAKE not particularly prominent and attached to the occiput by a relatively broad base. In contrast, desmognathine condyles are stalked , or pedunculate . They protrude from the occiput on relatively lengthy processes, thereby increasing the distance between the skull and the atlas. ModiJied atlas Dunn (1926), Noble (1927, 1931), Soler (1950), Wake (1966), Hinderstein (1971). The desmognathine atlas is modified in several ways. It is raised above the level of the trunk vertebrae and its centrum lies at a small angle relative to the longitudinal axis. The atlas has a large, raised, transverse ridge across the posterior end of the neural arch that receives the atlanto-mandibular ligaments (see below) on its anterior aspect. The condylar facets are enlarged and the odontoid processes greatly reduced. These do not fit into the foramen magnum as in other salamanders. Modified anterior trunk vertebrae Soler (1950), Wake (1966), Worthington dY Wake (1972), Hinderstein (1971). The anterior trunk vertebrae in desmognathines differ in two ways relative to other plethodontids. The first three trunk vertebrae are especially robust (at least in Desmognathus),having greater mass and width than more posterior vertebrae. Second, the first three to ten trunk vertebrae have accessory processes (pterygapophyses) arising from the dorsal surfaces of the postzygapophyses. Atlanto-mandibular ligament5 Dunn (1926), Noble (1927; 1931), Soler (1950), Wake (1966), Hinderstein (1971), Dalrymple et al. (1985), Larsen dY Beneski (1988). The most often described peculiarity of desmognathine cranial anatomy is the presence of stout ligaments that run from the posterodorsal transverse ridges of the atlas (see above), across the atlanto-occipital joint, to the coronoid processes of the mandible. The ligaments lie in deeply concave, longitudinal troughs in the parietal and otic bones. These parietal-otic troughs are also unique to desmognathines. They form a distinct posterolateral aspect of the parietal that is separated from the flat, smooth anterior portion by a ridge. The atlanto-mandibular ligament is associated with the superficial portion of the adductor rnandibulae anterior (the temporal muscle of earlier workers). The precise nature of this association has never been described, to our knowledge. Dunn (1926: 45) noted only that the atlas slip of the temporalis is tendinous in Desmognathus and Leurognathus , a description reiterated by Noble (1927: 6). Soler (1950: 469) wrote that fibres of the adductor encircle the ligament, whereas Hinderstein (1971: 471) reported the ligament enclosed by the muscle. It is probable that ancestrally the ligament was directly associated with the adductor muscle and was, in fact, its tendon, lacking any connection to the atlas. Such an incipient condition is retained in some non-desmognathine salamanders (Dubecq, 1925; Francis, 1934; Baird, 1951; Wake, 1966; Carol1 & Holmes, 1980). The brief descriptions noted above suggest that the atlanto-mandibular ligaments are anatomically and therefore, functionally, independent of any SALAMANDER FORM AND FUNCTION I45 surrounding musculature (see fig. 9 in Wake, 1966, and fig. 8 in Larsen & Beneski, 1988). The ligament is histologically typical of most tendons and ligaments in comprising collagen fibres almost entirely with no evidence of elastin (Dalrymple el al., 1985). Enlarged dorsal spinal muscles 3 Wake (1966), Hinderstein (1971), Worthington t Wake ( 1 9 7 2 ) . Particularly well-developed dorsal spinal muscles characterize desmognathine salamanders. These originate from the dorsal surfaces of the anterior trunk vertebrae, notably including the pterygapophyses described above. Insertion is on the back of the skull extending from the otic crests posteriorly to the parietal bone anterior to the parietal-otic trough. Enlarged quadrato-pectoralis muscles Pialt (1935), Soler (1950), Eaton (1937, 1957), Wake (1966), Hinderstein (1971). Worthington 9 Wake (1972). This muscle [called the gularis by Eaton (1937, 1957) and some others] originates on the pectoral fascia and skin of the gular fold. It inserts onto the quadrate and articular bones, and fascia of the depressor mandibulae and adductor musculature. In the resting animal, the bulge of this muscle on the ventrolateral aspect of the head and neck is so apparent that it creates what Piatt ( 1935: 222) referred to as the characteristic Desmognathus profile . The muscle is so-enlarged in all desmognathines. Hind limbs relatively larger than forelimbs Wake (1966). In all three desmognathine genera there is a disparity in size between the forelimbs and the hind limbs. This size disparity, reflected in limb length and to a lesser extent, limb robustness, is pronounced in Desmognathus and Leurognathus. In Phaeognathus, all four limbs are secondarily reduced, but the hind limbs remain larger than the forelimbs. The developmental basis of this disparity is complex; forelimb size is smaller than expected, but it is not certain that hind limb size is larger than expected (Wake, unpublished data). REVIEW OF PUTATIVE FUNCTION AND BIOLOGICAL ROLE An historical review of ideas relating to desmognathine cranial and cervical function and biological role is presented in Table 1. Note that the distinction between function and biological role is our own, none of the studies cited here differentiated between them explicitly. Hinderstein (1971) and Larsen & Beneski ( 1988) distinguished between purely functional hypotheses and presumed selective forces . The distinction is worth making because conclusions regarding biological role of anatomical features in the absence of experimental and/or natural history data are subjective, and as such, difficult to falsify. In contrast, purely functional conclusions regarding mechanical attributes of the system are more easily tested, either experimentally or theoretically. As a consequence, disagreements have arisen regarding the adaptive significance of desmognathine 146 K. SCHWENK AND D. B. WAKE TABLE An historical review of hypotheses relating t o t h e function and biological role of 1. desmognathine salamander specializations. Distinctions between function and biological role a r e mostly our own. C h a r a c t e r n u m b e r s refer t o t h e morphological characters listed a t t h e bottom of t h e T a b l e and discussed i n t h e text Source Dunn (1926) Characters ~~ Proposed functions(s) 1, Mandibular depression prevented 2. Skull raised to open mouth 3. Downward bending of head prevented 4. Anterior body stiffened Proposed biological role Burrowing 3, 4, 6 Noble (1927, 1931) 3, 4, 6 I . Mandibular depression limited 2. Skull raised to open mouth fully 3. Head strengthened Increase in head-raising muscle leverage 1. Mandibular depression limited 2. Skull raised to open mouth 3. Cranio-atlantal flexibility increased for skull elevation 4. Anterior body stiffened Burrowing Eaton (1957) Wake (1966) 3, 4, 6-8 None specified 1. Burrowing 2. Locomotion (streamlining) 1-9 Brodie (1978) Brodie ct al. (1989) Hinderstein (1971) None specified Increased bite force 1. Mandibular depression limited 2. Skull raised to open mouth fully 3. Cranio-atlantal flexibility increased 4. Mandible held against quadrate 5. Increased head weight Predator repulsion 1. Feeding 2. Burrowing I , 3-7 Worthington & Wake (1972) 3-8 1. Mandibular depression limited 2. Head used as wedge 3. Head dorsoflexion for substrate pushing Burrowing Dalrymple ct al. ( 1985) 1, 2, 6, 8 I . Increased bite force by means of 2. 3. 4. 5. Larsen & Beneski ( 1988) elastic energy storage and increased cranial inertia Increased bite force and jawlocking through cranial depression Jaws held firmly closed Head used as wedge Head dorsoflexion for substrate pushing 1. Feeding 2. Burrowing 6 I . Mandibular depression limited 2. Increased bite force I . Feeding (2. Burrowing) (3. Predator repulsion) (4. Courtship) Characters: 1. Robust skull and mandible 2. Flat, wedgelike head profile 3. Stalked occipital condyles 4. Modified atlas 5. Modified anterior trunk vertebrae 6. 7. 8. 9. Atlanto-mandibular ligaments Enlarged dorsal spinal muscles Enlarged quadrate-pectoralis muscles Relatively larger hind limbs SALAMANDER FORM AND FUNCTION 147 specialization even in the light of identical functional conclusions. For example, nearly every study has concluded that mandibular depression is to some extent limited in desmognathines, but these studies are almost equally divided between burrowing and feeding as the biological role of choice for such a mechanism. Cope ( 1889) called attention to several desmognathine cranial specializations, noting the peculiar morphology of the adductor musculture and its tendons, the occipital condyles, and the strengthened skull bones. He believed that these features enabled Desmognathus to burrow among stones and in earth more readily than the species of other genera (p. 196). Dunn (1926) noted the presence of the atlanto-mandibular ligaments, modified atlas and stalked condyles, and considered them diagnostic of Desmognathinae. He regarded them as a cohesive suite of characters and was the first to relate them to a specific functional hypothesis, as well as a biological role: Thus the lower jaw is immovably attached to the atlas, and the mouth is opened by raising the skull. The function of this arrangement, with which is connected the stalked condyles, and collared atlas of these forms, is to prevent the downward bending of the head, and to stiffen the anterior part of the body (p. 45), so that it can be forced under things (p. 59). Noble (1927, 1931) agreed with Dunn (1926) except that he found that some mandibular depression was possible. In contrast, Eaton (1957) rather vigorously disagreed with both Dunn and Noble. He dissected a single Desmognathus quadramaculatus and found that the adductor muscle was only partly tendinous , though his description suggests that his specimen might have been a juvenile. In any case, Eaton s (1957) electrical stimulation of the superficial muscles in an anaesthetized individual led him to dismiss Dunn s (1926) hypothesis of an immobile mandible; he further concluded that the mechanism of head elevation was identical in non-desmognathine salamanders and believed that desmognathines differed from other species in certain attributes only in degree. T o these differences he ascribed no particular function or biological role other than an increase in leverage for head-elevating muscles. Wake (1966) reviewed the morphology of the Desmognathinae and other plethodontids and described all of the putative spcializations noted in this paper. For the most part Wake (1966: 47) agreed with Dunn (1926) in characterizing desmognathines as a group of salamanders with a unique mouth opening mechansism by means of which the mandibles are held relatively rigid and the skulls proper are raised [as well as having] extensive skeletal and muscular modifications related to functional changes in mouth opening mechanics . However, Wake (1966: 57) also accepted Noble s (1927, 1931) findings in emphazing that, The mandibles are far from immobile in desmognathines . He further agreed that the anterior part of the body would be stiffened, but that the atlanto-occipital joint is more flexible than in plethodontines. Wake ( 1966) did not attribute the mechanics of mouth opening to any particular biological role; however, he regarded the wedge-shaped, robust desmognathine skull and mandible, hypertrophied dorsal spinal muscles, and relatively larger hind limbs as adaptations for burrowing. Hinderstein (1971) more or less reiterated Wake s (1966) functional conclusions, although he stated explicitly several hypotheses regarding the function of the atlanto-mandibular ligaments (Table 1). He departed from 148 K. SCHWENK AND D. B. WAKE Wake s (1966) views, however, in concluding that the unique desmognathine mechanism is an adaptation for feeding. Specifically, he envisioned these salamanders feeding on the bottoms of fast-moving streams, anchored there by their massive crania. Resting with their mandibles on the stream bottom would then necessitate mouth opening by cranial elevation. He ascribed a secondary role for this mechanism to burrowing. Worthington & Wake ( 1972: 275-276) succinctly described a specific mechanism whereby the desmognathine morphology might be used in burrowing: A stout ligament extends from the atlas to the mandible, restricting the downward motion of the mandible. Strong throat muscles pull the head down to the point at which the temporal [atlanto-mandibular] ligament becomes taut, and the head, bent sharply downward, is used as a wedge by the organisms, to allow them to squeeze under rocks and into burrows. In order to function effectively, the head must be raised, and this task is accomplished mainly by the dorsal spinal muscles, which are greatly hypertrophied in desmognathine salamanders . These conclusions were based on observations of living desmognathines. Two recent studies have examined experimentally the role of the atlantomandibular ligaments. Dalrymple el al. (1985) and Larsen & Beneski (1988) both studied feeding in Desmognathus. Dalrymple et al. (1985) discovered that immediately after lingual ingestion, the salamander often assumed a so-called head-tuck posture in which the entire head was ventroflexed with the mouth shut. Occurrence of this behaviour was significantly correlated with the size of the prey item and its position in the mouth after ingestion. Salamanders were more likely to head-tuck if prey protruded from the jaws after ingestion. Electromyographic recordings showed the quadrato-pectoralis muscles to be active during head-tucking. They found that the head was elevated during mouth opening by the dorsal spinal muscles and presented circumstantial evidence suggesting a large bite force in Desmognathus. They concluded that the desmognathine jaw mechanism is specialized as a kinetic-inertial mechanism, sensu Olson (1961), arguing that such a mechanism would confer a more powerful bite owing to two factors: (1) the atlanto-mandibular ligaments, stretched during jaw opening, would store elastic strain energy; this energy would be released during jaw closure by elastic recoil of the ligaments; (2) increased mass of the skull and mandible. By increasing mass and acceleration of the jaws relative to other salamanders, bite force would be increased ( F = ma). A secondary conclusion was that the atlanto-mandibular ligaments would also be stretched during head-tucking and that these would help to keep the jaws closed while applying static pressure to large or active prey. These functional conclusions led Dalrymple et al. (1985) to propose that desmognathines are adapted to eat relatively larger prey, but they noted anecdotally that head-tucking, followed by skull dorsoflexion, also occurs during burrowing. Therefore, they did not rule out burrowing as a factor in the evolution of this mechanism. Larsen & Beneski s (1988) high-speed cine analysis of feeding in Desmognathus showed that mandibular depression is restricted, but not stopped, by the atlantomandibular ligaments. They also showed that cranial elevation contributes significantly to gape during mouth opening, but pointed out that this is true in other, non-desmognathine salamanders as well. They concluded that cranial SALAMANDER FORM AND FUNCTION I49 dorsoflexion during mouth opening is independent of any peculiar morphological adaptations (p. 1309) and concurred with the principal conclusions of Dalrymple et al. (1985). Larsen & Beneski (1988) suggested that functional aspects of desmognathine head and neck mechanics differ only in degree, and possibly efficiency, from other salamanders, agreeing in this sense with Eaton (1957). They went on to note that no consensus had emerged regarding the biological role of the desmognathine specializations, but suggested four possibilities: feeding, burrowing, predator repulsion and courting. They concluded that the primary selective force shaping this structural complex was that it allowed an increase in bite force without a concomitant increase in head and body size. Improvements in the other biological roles were viewed by these authors as secondary to that of feeding. PREY PROCESSING IN LEUROGNATHUS MARMORATUS Background and method5 The genus Leurognathus comprises a single species, L. marmoratus. It is a fully aquatic salamander inhabiting sandy and rocky bottom pools of Appalachian Mountain streams in the south-eastern United States (Pope, 1924, 1928; Dunn, 1926; Pope & Hairston, 1947; Hairston, 1949; Martof, 1962). It feeds primarily on aquatic insects and the aquatic larvae of terrestrial insects (Martof & Scott, 1957). Remarkably, Leurognathus mannoratus utilizes a fundamentally terrestrial feeding mechanism (tongue projection) when feeding under water (Schwenk & Wake, 1988, and in preparation). We observed feeding in 22 captive specimens over a period of several months. Specimens fed readily on waxmoth larvae (Galleria mellonella) dropped into the water and less readily on earthworm segments. Two specimens were induced to feed in a small, plexiglass filming chamber filled with water. These were filmed at 50 frames per second (measured) with a Bolex 16 mm cine camera and synchronized stroboscopic illumination (Helmuth-Chadwick) . Nineteen feeding sequences were filmed. Of these, seven sequences were complete, including ingestion, intra-oral transport, and deglutition [see Bramble & Wake (1985) and Schwenk & Throckmorton ( 1989) for discussions of vertebrate feeding stages]. Ten additional sequences included ingestion and at least the first part of intraoral transport, and the remaining two sequences included lingual ingestion only. Films were analysed frame by frame on a Vanguard motion analyser. Still 35 mm photographs were produced from individual 16 mm cine frames (Fig. 1); tracings and measurements were made directly from the motion analyser screen. Prty processing in Leurognathus Lingual projection and prey capture (ingestion) in Leurognathus is similar to that reported for Desmognathus Larsen & Beneski, 1988) with the exception that the entire feeding sequence occurs under water (Schwenk & Wake, 1988). Importantly, the feeding cycles of Leurognathus and Desmognathus are similar in most essentials to feeding in a variety of other lingual-feeding salamanders (e.g. Larsen et al., 1989; Reilly & Lauder, 1989, 1990; Miller & Larsen, 1990). I50 K. SCHWENK AND D. B. WAKE SALAMANDER FORM AND FUNCTION 151 However, one conspicuous difference was found in the feeding cycle of Leurognathus as compared with other salamanders: the occurrence of sharp, rapid cranial ventroflexion(s) immediately following ingestion (Figs 1, 2). This behaviour appears to be identical to that reported in Desmognathus by Dalrymple et al. (1985) and called by them head-tucking (see above). The behaviour involves strong ventroflexion of the cranium at the atlanto-occipital joint anywhere from one to 13 times after prey capture, but always preceding intraoral transport proper. In most cases the mouth is opened wide before the head tuck, principally by cranial elevation. The head tuck is then initiated as the jaws close in a rapid bite. In some cases, a bite does not precede the head-tuck (e.g. Fig. 2). In these instances, the jaws are only slightly parted at the start of the head-tuck owing to the presence of the prey item between the upper and lower tooth rows (see below). As the head-tuck progresses, gape is decreased until full jaw closure is achieved at maximum cranial ventroflexion. The head is then elevated and as this occurs the jaws part again, seemingly from elastic recoil of the body of the prey item. The head-tucking sequence, with or without a bite, is then usually repeated. Head-tucking occurred in nine of 16 analysable feeding sequences and was significantly correlated with the presence of the prey item between the teeth (f = 8.96; P< 0.003; df = 1); indeed, in only one case did a salamander fail to head-tuck when the prey item came to be held between the teeth after ingestion. In that sequence the waxworm was transported and swallowed immediately following capture. In two cases, the waxworm was initially held within the mouth but was promptly transferred by the tongue to a position between upper and lower anterior tooth rows. Only then was head-tucking initiated. In six other cases the waxworm was brought wholly within the mouth after prehension and transported for swallowing directly, without further manipulation and without head-tucking. Following head-tucking, a series of hyolingual transport movements is initiated [never inertial transport as stated by Dalrymple el al. (1985); see also Larsen & Beneski (1988)l and the prey is swallowed. DISCUSSION We believe that most studies have been hindered in their functional analyses to a greater or lesser extent because they have assumed, a priori, that all Figure 1. Leurognathus marmoratus, a desmognathine plethodontid salamander, exhibiting a bite and single head-tuck following ingestion of a waxmoth larva. Photos from consecutive frames of a 16 mm cine film shot at a measured rate of 50 fps. Sequence filmed with the salamander completely submerged in water. The salamander is motionless at the start of the sequence (frame 1). Note that the larva is held between the tooth rows and that the jaws are held apart by the thickness ofits body. Also note the nearly flat resting angle between the cervical and cranial axes, indicating that the head is only slightly bent at the atlanto-occiptal ( A - 0 ) joint). During the bite, maximum gape is achieved by elevation of the cranium and depression of the mandible while the worm remains adherent to the tongue (frames 2 and 3). Jaw closure occurs in frames 4 and 5. At the moment of upper jaw contact with the prey item (frame 5), cranial ventroflexion (VF) is initiated at the A - 0 joint. As VF proceeds, the A - 0 angle is further decreased by dorsal bending of the cervical vertebral axis until maximum cranial V F (minimum A - 0 ) angle is achieved by frame 7. Subsequently (frames 8-10 and beyond), V F slowly decreased until the resting position is again attained. Note that at the beginning of the head-tuck sequence (frame 6), upper and lower tooth rows are in contact and the worm is effectively crushed. Additional V F serves to increase the force of this bite, a static-pressure system (see text). Head-tucks are usually repeated, with or without a preceding bite. 152 K. SCHWENK AND D. B. WAKE 200 190 180 ~ ' ~ ' " " " ~ ' " ~ ' ~ ' " ~ ' " " " ~ ' 170 160 150 140 130 120 110 - -0. 10 . , , 30 20 40 . I r I I I I . I . I . 1 I . , 50 60 70 80 90 100 110 120 130 140 150 Cine Frame Number Figure 2. Plot illustrating changes in the angle between the vertebral axis and the cranium (atlanto-occiptal angle, 0, illustrated above the plot) for an ingestion (prey capture) sequence and six subsequent head-tucks in Leurognalhus marmoralus feeding underwater. A - 0 angle was measured manually from tracings madc on a motion analyser screen from 16 mm cine film. Frame rate measured at 50 fps. A decrease in 0 represents cranial ventroflexion, or head-tucking. Note the rapidity of the head tucks (c. 40 ms duration) and the relatively slower decay to the rest position. I t is likely that return to the rest position is caused by elastic recoil of the atlanto-mandibular ligaments, and not muscular action, although we have no direct evidence of this. Also note that head tucks occur in the absence of inital jaw opening, with the upper and lower jaws already in contact. In these instances a kinetic-inertial mechanism cannot possibly be operating (see text). Slight deviations evident in tJ undoubtedly reflect measurement error, which was determined to be c. f I". desmognathine morphological modifications are functionally related in some way. We suggest that there are at least two forces shaping desmognathine head and neck morphology. Thus we believe that four characters must be considered independent of the others in a purely functional sense: stalked occipital condyles; modified atlas; atlanto-mandibular ligaments; and hypertrophied quadratopectoralis musculture. These characters can be explained as a functional complex operating in the context of feeding. The remaining five characters [robust (massive) skull and mandible; flat, wedgelike head profile; SALAMANDER FORM AND FUNCTION I53 hypertrophied dorsal spinal muscles; modified anterior trunk vertebrae; relatively and larger hind limbs] probably act in concert with this functional complex, but in the context of head-first burrowing. Desmognathine specialization and feeding Early hypotheses of mandibular fixation or significant restriction in desmognathines (Table 1) have been falsified by recent cinegraphic studies (Larsen & Beneski, 1988; Larsen el al., 1989; this study). Dunn s (1926) suggestion that downward bending of the head would be prevented by the atlanto-mandibular ligaments has been rejected. Indeed, cranial ventroflexion is a striking feature of the desmognathine feeding mechanism (see below). In contrast to previous suggestions, several studies have now shown that cranial elevation during mouth opening is ubiquitous among salamanders, not unique to desmognathines and, therefore, cannot be associated with desmognathine specialization (Larsen & Guthrie, 1975; Roth, 1976; Lombard & Wake, 1977; Dockx & de Vree, 1986; Larsen el al. 1989). A number of workers have suggested adaptation for increased bite force in desmognathines relative to other salamanders, but evidence is circumstantial: ( 1 ) the jaws produce an audible snap when closed (Brodie, 1978); (2) the jaws are used defensively against formidable vertebrate predators such as shrews and snakes, and seem to lock shut (Brodie, 1978; Brodie, Dowdey & Anthony, 1989); (3) earthworm prey are often cut in half when captured (Dalrymple et al., 1985). There are no direct measurements of bite force in desmognathines or any other salamander as far as we know, therefore conclusions regarding relative bite force among species are highly speculative. Dalrymple el al. (1985) proposed a kinetic-inertial (K-I) model to account for this putatively larger bite force (see Review, above). Their model is based on the discussion of Olson (1961: 209) who characterized K-I mechanisms in the following way: The effective action of the jaw and teeth depends primarily upon forces developed by rapid motion of the lower jaws relative to the upper. Once motion has been initiated, it is the velocity and mass of the jaws that are effective. The j a w s at rest tend to exert little force and in or near occlusal position the forces serve to do little more than holdposition [italics added]. Support for the K-I model is weak, however. For example, in the discussion of bite force above, only point one is consistent with a K-I mechanism. Point two and probably point three, relate to bite force generated after occlusion. Comparative and biomechanical considerations further undermine the case for a K-I system being especially welldeveloped in desmognathines. Many generalized, non-desmognathine salamanders possess a jaw mechanism that should function as well or better in elastic energy storage during jaw opening. In these species the mandible and atlas are connected by the levator (adductor) mandibulae anterior (superficial portion) and its associated tendon (Baird, 1951; Hinderstein, 1971). This is the muscle through which the atlanto-mandibular ligament runs in desmognathines and the tendon from which, presumably, the ligament arose ancestrally. For example, in the plethodontine genus Pseudoeurycea, the muscle is a heavy, cordlike bundle of fibers, [originating] from a short heavy tendon attached to the neural spine of the atlas (Baird, 1951: 243). Typically, vertebrate mechanisms using elastic energy storage are just such muscle-tendon systems that exploit the I54 K. SCHWENK AND D. B. WAKE mechanical properties of both tissues, including the important series elastic component of muscle (e.g. Cavagna, 1970; Cavagna & Citterio, 1974; Cavagna, Heglund & Taylor, 1977). Therefore, isolation of the tendon from the muscle, as apparently occurred in the evolution of the atlanto-mandibular ligament, would confer no advantage to the elastic strain energy storage mechanism already present ancestrally and may, in fact, have limited it. Such isolation would eliminate the elastic muscle component from the atlanto-mandibular linkage, leaving only the highly inelastic ligament. The stiffness of ligament, resulting from the high elastic (Young s) modulus of collagenous tissue (Wainwright et al., 1976), suits it to its usual role in force transmission or movement restriction. We do not disallow the possibility of the desmognathine mechanism functioning in elastic energy storage. Indeed, the inelastic nature of the atlanto-mandibular ligaments might enable them to store energy more efficiently over very small excursion distances and there is no denying the important role of tendon elasticity in vertebrate locomotion (e.g. Alexander, 1984; Dimery, Alexander & Ker, 1986). However, we point out the unusual nature of an isolated, collagenous ligament acting in elastic strain energy storage and emphasize the role of the atlanto-mandibular ligaments in transmitting force to the mandible, as it is this function that seems to characterize uniquely the desmognathine salamanders (discussed below). Our study of Leurognathus, as well as observations of Dalrymple el al. (1985) on Desmognathus, indicate that the one unique behavioural trait characteristic of desmognathine feeding is that of cranial ventroflexion, or head-tucking. It is therefore reasonable to conclude that it is this behaviour that is associated with at least some of the head and neck specializations exhibited by desmognathines. If one accepts that cranial ventroflexion with the mouth closed is the salient feature of desmognathine behaviour, and not cranial dorsoflexion in order to open the mouth (as previously supposed), then the first four morphological characters noted above are easily interpreted: stalked condyles increase the degree of cranial ventroflexion possible at the atlanto-occipial joint by allowing the skull to swing through a greater arc. The force for this ventroflexion is provided by the massive quadrato-pectoralis musculature (Dalrymple et al., 1985). The hypertrophy of these muscles relative to plethodontines reflects their potentially greater force production. Greater force production is also enhanced by the stalked condyles which increase the quadrato-pectoralis moment arm (i.e. its mechanical advantage (Fig. 3). As the skull is ventroflexed at the atlantooccipital joint the atlanto-mandibular ligaments become taut. The atlas is modified to receive the stout ligaments and to resist the potentially large reaction forces generated at the atlanto-occipital joint. Most importantly, however, the force required to tense the ligaments is provided by the powerful quadratopectoralis musculature and not the relatively feeble depressor mandibulae muscles. Recruitment of these large axial muscles would provide much more power to act on the ligaments than would be possible through mandibular depression (or cranial elevation with mandibular stabilization). During cranial ventroflexion, the mandible would be pulled into occlusion with the upper jaw. Thus jaw closure would be accomplished by active depression of the cranium and passive elevation (relative to the cranium) of the mandible through tension in the ligaments. Once the cranium is depressed beyond the rest position, the ligaments must become increasingly taut, even with the tooth rows of upper and SALAMANDER FORM AND FUNCTION I55 C Figure 3. Schematic illustration of a desmognathine salamander skull (lateral view) showing changes in mechanical advantage obtaining from the evolution of pedunculate occiptal condyles. A, Pedunculate condyles would increase the mechanical advantage of reaction forces (F) operating against the skull during burrowing. Such forces are divisible into horizontal (h) and vertical (v) components. The vertical component of the force would tend to bend the salamander's head away from the axis, an action highly disadvantageous during burrowing. Its lever arm (2) is greater than would be present with plesiomorphic, sessile condyles (1). Hence pedunculate condyles tend to increase atlanto-occipital flexibility. B, In contrast, a similar increase in the length of the lever arm for the vertical component of the quadrato-pectoralis muscle (Q-P) (2 vs I ) increases the efficacy of this muscle in causing cranial ventroflexion during head-tucking, resulting in increased bite force after jaw closure and the promotion of an effective static-pressure system. Mechanical advantage of the dorsal spinal muscles is similarly improved (see text). lower jaws in full occlusion. The collagenous ligaments are ideally suited to transmit the full force of the quadrato-pectoralis musculature to the mandible. In this way, an extremely powerful bite force can be generated by recruitment of hypertrophied axial muscles as opposed to the relatively smaller and weaker adductor mandibulae musculature. Our conclusions differ from those of Dalrymple et al. (1985) and Larsen & Beneski (1988) mostly in emphasis; these workers stressed the role of the desmognathine jaw mechanism in producing an effective kinetic-inertial system. While a kinetic-inertial jaw mechanism is to some extent present in the Desmognathinae as in most tetrapods (see above), we do not believe that its development is significantly greater than in plethodontines or other salamanders and, therefore, cannot account for the unusual morphological attributes of the desmognathines. Rather, we believe that this system is an extreme expression of I56 K. SCHWENK AND D. B. WAKE the alternate mechanism in Olson s (1961: 209) dichotomy, a static-pressure system, in which, The primary action results from pressures exerted by the jaws when nearb closed and initially not in motion [italics added]. It is the development of powerful occlusal forces once the jaws are closed and no longer moving relative to one another that uniquely characterizes the remarkable desmognathine mechanism. Indeed, head-tucking in Leurognathus sometimes occurs in the absence of mouth opening (Fig. 2). Dalrymple el al. (1985) noted this function of head-tucking, but considered it secondary to the K-I system. Dalrymple et al. (1985) and this study have established that desmognathines are unique in the way that they handle prey once captured, immobilizing and possibly killing their prey with a specialized static-pressure jaw mechanism utilizing cranial ventroflexion. There is a precise form-function correlation between the observed behaviour and unique attributes of desmognathine anatomy and mechanical analysis suggests that they are functionally linked. Therefore, we conclude that feeding has been a significant determinant of cranial and cervical specialization in desmognathine salamanders. Desmognathine specialization and burrowing While cranial ventroflexion and a powerful static-pressure jaw mechanism account for four desmognathine characters, five others remain unexplained: heavily ossified skull and mandible; flat, wedgelike head profile; hypertrophied dorsal spinal muscles; modified anterior trunk vertebrae; and relatively larger hind limbs. Below, we discuss the biomechanics of head-first burrowing, consider desmognathine morphology in light of this discussion, and note the observational and ecological evidence in support of burrowing as a factor in desmognathine evolution. Gans (1960, 1974) described three components of head-first burrowing: substrate entry, tunnel extension and tunnel widening. Animals might exhibit all, none, or a subset of these. For example, salamanders might exploit existing crevices and holes in the ground without creating their own. However, existing crevices that are too narrow to permit entry might be widened and possibly extended, or a salamander might create a wholly new tunnel. Exploitation of existing crevices requires no particular specialization, but active burrowing is likely to require some form of morphological and functional modification. Differences among species might reflect differences in the manner and extent to which each activity is performed. Crevice widening requires the ability to move the substrate to the sides, compacting it and enlarging the opening. Amphisbaenians accomplish this by forcefully moving the snout and head dorsad (spade-snouted forms) or side-toside (keel-snouted forms; Gans, 1974). In most cases they generate powerful flexion of the head through use of segmental axial muscles acting on the cranium through a tendon. Tunnel extension and initial substrate entry entail the creation of an opening in solid substrate. In amphisbaenians and other head-first burrowers this is usually accomplished by ramming. Ramming pushes the anterior end of the cranium into the substrate, either displacing it to the sides or compacting it anteriorly. Compaction would be detrimental to subsequent penetrations, hence a skull form that maximizes displacement is advantageous. Gans (1974) noted that a long, slender, pointed cone would be maximally efficient for substrate SALAMANDER FORM AND FUNCTION I57 penetration, but would be likely to fail; shaping the snout like a wedge is one strategy for strengthening the snout while retaining most of the advantages of a cone. This is the strategy adopted by most amphisbaenians. Ramming and pushing the substrate must also place significant stress on the skull. Head-first burrowers generally exhibit robust skulls characterized by increased ossification and greater suture complexity. This is true in gymnophione amphibians (Wake & Hanken, 1982; Wake, 1993), amphisbaenians (Gans, 1960, 1974), uropeltid snakes (Gans, 1973), and some burrowing mammals (Hildebrand, 1985; Wake, 1993). Gans (1960) pointed out that the force required for a given degree of tunnel penetration is proportional to the cross-sectional area of the burrower. Amphisbaenians reduce head and body diameter while maintaining forceful cranial movement by recruiting muscle fibres distributed over several body segments. Finally, Gans (1974) noted that many amphisbaenians assume a snoutdepressed posture when penetrating the substrate. He pointed out that this functionally countersinks the lower jaw so that it cannot be forced open by shear from the substrate. It also serves to concentrate reaction forces to the dorsal surface of the snout which, in amphisbaenians, is variously modified to resist such forces. Furthermore, in such forms further snout depression, resulting from these reaction forces, is resisted by a nuchal tendon joining the dorsal axial skeleton to the cranium. Burrowing is a compelling explanation for the evolution of desmognathine form because it plausibly relates and explains the remaining five features noted at the outset of this section. The heavily ossified skull and mandible, with their strongly articulated sutures, resist stresses caused by ramming or pushing into crevices. The flat, wedgelike head profile displaces the substrate, minimizing its anterior compaction while maintaining skull strength. T h e hypertrophied dorsal spinal musculature acts in conjunction with the hypertrophied quadratopectoralis muscles to provide powerful dorsoventral flexion of the head for substrate compaction during crevice widening. We note that this behaviour exploits the stalked occipital condyles, which increase the mechanical advantage of these muscles (Fig. 3). The condyles also increase the mechanical advantage of substrate reaction forces, but these could be resisted by the hypertrophied, antagonistic cervical musculature and by assumption of a snout-depressed posture. This posture would serve to accentuate the wedgelike head profile, placing the atlanto-mandibular ligaments in tension and concentrating bending forces onto the top of the snout. Further snout depression would be resisted by the tensed ligaments which would also lock the mandible closed and prevent mouth opening from substrate shear. Furthermore, use of posteriorly placed axial musculature distributed over more than one body segment not only provides more force through recruitment or more fibres, but serves to minimize head diameter and, therefore, the energy necessary for substrate penetration. Finally, the relatively larger hind limbs of desmognathines provide greater thrust for initial ramming, wedging or crevice widening while the smaller forelimbs maintain a reduced cross-sectional area anteriorly. Thrust might also be increased by rapidly straightening an initially flexed, elongate body. Observational evidence supports our proposed mechanism of head-first burrowing in desmognathines. A number of early workers, including Cope (1889) and Dunn (1926) knew the propensity of Desrnognathus species to wedge I58 K. SCHWENK AND D. 8.WAKE themselves under objects such as stones. Worthington & Wake (1972: 275-276) observed that Desmognathus used head-tucking to allow them to squeeze under rocks and into burrows and noted that this is followed by head elevation. Similarly, Dalrymple et al. (1985: 257) commented that desmognathines used head-tucking to push substrate with their heavy, flattened skulls in burrowing . . . and then the entire head is dorsoflexed as a wedge for pushing aside substrate . Brandon ( 1965: 70) performed simple experiments with Phaeognathus and found that, when attempting to penetrate dense soil, his specimens repeatedly lifted the posterior part of the head and pushed against the soil with the snout , a clear description of head-tucking. These observations indicate that desmognathines are, indeed, burrowers that utilize head-tucking and dorsoventral cranial flexion, as outlined above. Unlike feeding specialization, which seems to be consistent among all desmognathine species, burrowing ability is likely to vary. This variability probably relates to differences in habitat use among genera. For example, Leurognathus marmoratus is a fully aquatic species whose burrowing behaviour is restricted to wedging itself under stream-bottom rocks. Martof (1962) noted that Leurognathus prefers rocky to sandy stream bottoms. In contrast, Phaeognathus hubrichti is fully terrestrial and is known to burrow actively in soil (Brandon, 1965; Wake, unpublished data). Species of Desmognathus are intermediate in habitat use, but are most often described as wedgers , like Leurognathus (Dunn, 1926; Noble, 1927; Wake, 1966; Worthington & Wake, 1972). Because the mechanical requirements of wedging under rocks are probably somewhat different from tunnelling through soil, we might expect intergeneric differences among desmognathines in the degree and nature of specializations relating to burrowing. The preliminary observations noted above suggest that all genera are subjected to ramming forces resulting from wedging (Leurognathus, Desmognathus) or substrate entry (Phaeognathus, Desmognathus?). Brandon s ( 1965) observations confirm that, given the right conditions (loose enough soil or an initial starting crevice), Phaeognathus is capable of creating its own tunnel. Indeed, Phaeognathus is associated with burrows in the field (e.g. Valentine, 1963; Huheey, 1964; Dodd, 1988) and there is morphological evidence suggesting that it is particularly specialized for burrowing (Wake & Schwenk, unpublished). While feeding and burrowing seem to be the principal biological roles for desmognathine specializations, other, secondary roles are possible. The staticpressure jaw mechanism, for example, benefits the animal in any situation in which a powerful bite would be of use, as in defence against predators (Brodie, 1978; Brodie el al., 1989). The fact that more generalized salamanders lacking such specialization have also evolved powerful bites used in antipredator behaviour (e.g. Noble, 1931; Stebbins, 1951; Wake, 1963; Nussbaum, 1976; Brodie, 1977; Feder & Arnold, 1982) suggests that this was not a significant factor in the origin of this mechanism. One can imagine other scenarios in which the desmognathine jaw mechanism might be useful (intraspecific courtship and agonistic behaviour, for example; see Arnold & Houck, 1982, for discussion of the former), but there is presently no evidence for these. Evolutionary conclusions Desmognathines share a suite of uniquely derived characters that are explicable as adaptations serving the complementary roles of feeding and head- SALAMANDER FORM AND FUNCTION I59 first burrowing. In order to discern evolutionary patterns among these characters it is necessary to consider them in light of a phylogenetic hypothesis. However, salamander relationships have long been controversial (Hecht & Edwards, 1977; Duellman & Trueb, 1986), and new molecular evidence (Larson, 1991 ) fails to corroborate hypotheses based on morphology, in particular the generally accepted view that Ambystomatidae is the sister group of Plethodontidae. Nor is a robust hypothesis of desmognathine species relationships available at this time. Therefore, interpretation of desmognathine synapomorphies in an historical context is difficult. Nonetheless, some generalizations can be made. Plethodontine plethodontids are morphologically more similar to any potential outgroups, including ambystomatids, than are desmognathines. Use of the head for crevice-widening and burrowing in soil is widespread among plethodontines and ambystomatids (Heatwole, 1960, and references therein), and is therefore likely to be plesiomorphic. However, the desmognathines appear to be far more proficient in these functions than other taxa and they are particularly effective in wedging head-first into the bottoms and sides of rocky streams. The morphological features enabling these behaviours, however, also contribute directly to the unique desmognathine mechanism of immobilizing and processing prey, once captured. Central to both roles is the presence of the atlanto-mandibular ligaments, which may be viewed as a key innovation in desmognathine evolution. However, in the absence of a robust phylogenetic hypothesis and any intermediate form, we are unable to argue convincingly for the primacy of either feeding or burrowing in the adaptive evolution of desmognathine salamanders. By key innovation we mean a trait that allowed desmognathines to exploit a new adaptive zone relative to their ancestors (sensu Baum & Larson, 1991). Although Wake ( 1966) argued that desmognathines retained the ancestral adaptive zone, new evidence, including use of the tongue in feeding underwater (Schwenk & Wake, 1988) and the several ways in which head mobility has been exapted during desmognathine phylogenesis (this paper), argues for a secondary invasion of stream habitats by adult desmognathines. Beachy & Bruce (1992) argued convincingly that lunglessness evolved as a rheotropic adaptation in stream larvae. In order to test this hypothesis it will be necessary to use a comparative, phylogenetic methodology, such as that advocated by Baum & Larson ( 1991 ) . The morphological specializations we have described are variably expressed among desmognathine species; mapping these features onto a cladogram might elucidate the historical sequence of character transformations that have led to the highly specialized feeding and locomotor systems of desmognathine salamanders. Such an analysis awaits a well-corroborated phylogeny of Desmognathinae. ACKNOWLEDGEMENTS We are very grateful to Wayne Van Devender for providing the Leurognathus used in this study, Keith Condon for help with much of the filming, and Fuzz Crompton for logistic support. Additional support was provided by Karen Hiiemae and the Department of Oral Anatomy, University of Illinois at Chicago. K.S. thanks Beth Brainerd for stimulating discussions about 160 K. SCHWENK AND D. B. WAKE biomechanics. Research was supported by NIH NIDR 5F32DE05467 and a grant from the University of Connecticut Research Foundation to K.S. REFERENCES Alexander R. McN. 1984. Elastic energy stores in running vertebrates. American zoologist 24: 85-94. Arnold SJ, Houck LD. 1982. Courtship pheromones: evolution by natural and sexual selection. In: Nitecki MH, ed. Biochemical aspects o evolutionary biology. Chicago: University of Chicago Press, 173--21I . f Baird IL. 1951. An anatomical study of certain salamanders of the genus Pseudocutyea. Universily o Kansas f Science Bulletin 34: 22 1-265. Baum DA, Larmon A. 1991. Adaptation reviewed: a phylogenetic methodology for studying character 0 macroevolution. $ystematic <oology 4 : 1-18. Beachy CK, Bruce RC. 1992. Lunglessness in plethodontid salamanders is consistent with the hypothesis of mountain stream origin: a response to Ruben and Boucot. American Naturalist 139: 839-847. Bock WJ, von Wahlert C. 1965. Adaptation and the form-function complex. Evolution 19: 269-299. Bruuble DM, Wake DB. 1985. Feeding mechanisms of lower tetrapods. In: Hildebrand M, Bramble DM, Liem K F & Wake DB, eds. Functional vertebrate morphology. Cambridge, Mass.: Harvard University Press, 230-261. Brandon RA. 1965. Morphological variation and ecology of the salamander Phacognathus hubrichti. Copeia 1965: 67-71. Brodie ED Jr. 1977. Salamander antipredator postures. Copeia 1977: 523-535. Brodie ED Jr. 1978. Biting and vocalization as antipredator mechanisms in terrestrial salamanders. Copeia 1978: 127-129. Brodie ED Jr, h w d c y TC, Anthony CD. 1989. Salamander antipredator strategies against snake attack: biting by Desmognathus. Herpctologica 45: 167- I7 I . Carroll, RL, Holmem R. 1980. The skull and jaw musculature as guides to the ancestry of salamanders. {oological Journal o the Linnean Sociep 68: 1-40, f Cavrgna CA. 1970. The series elastic component of frog gastrocnemius. Journal o Physiology 206: 257-262. f Cavrpa CA, Cittcrio C. 1974. Effect of stretching on the elastic characteristics and the contractilc component of frog striated muscle. Journal o Physiology 239: 1-14. f Cavapa CA, Heglund NC, Taylor CR. 1977. Mechanical work in terrestrial locomotion: two basic mechanisms for minimizing energy expenditure. American Journal o Physiology 233: R243-R261. f Cope ED. 1869. A review of the species of the Plethcdontidae and Desmognathidae. Proceedings o the Academy o f f Natural Sciences, Philadebhia 1869: 93-1 18. Cope ED. 1689. The Batrachia of North America. Bulletin o thc United States National Museum 34: 1-525. f Dalrymple CH,Juterboek JE,La Volley AL. 1985. Function of the atlanto-mandibular ligaments of desmognathine salamanders. Copcia 1985: 254-257. D i m c r y NJ, Alexander R McN, Kcr RF. 1986. Elastic extension of leg tendons in the locomotion of horses (Equus caballus). Journal o <oology, London 2101 415-425. f DOCLX P, De Vree F. 1986. Prey capture and intra-oral food transport in terrestrial salamanders. In: Rocek Z, ed. Studies in herpetology. Prague: Charles University, 521-524. Dodd CK Jr. 1988. A re-examination of the status of the Red Hills salamander, Phaeognathus hubrichti. Report to the U S . Fish and Wildlifc Service. Dubccq J. 1925. Constitution du muscle temporal chez les Amphibiens Urodeles: signification morphologique de ce muscle. Comptes Rendus de la Stances Sociltl de Biologie 93: 1523. h e l l n u n WE, Trueb L. 1986. Biology o amphibians. New York: McCraw-Hill Book Co. f Dunn ER. 1926. The salamanders o the farnib Plethodontidae. Northhampton, Mass.: Smith College. f Eaton TH Jr. 1937. The gularis muscle in Urodela. Journal o Morphology 6ot 317-324. f Eaton TH Jr. 1957. Motor areas and superficial muscle action in salamanders. Journal o the Elisha Mitchell f Scient$c Socic& 73: 1- 10. F e d a ME, Arnold SJ. 1982. Anaerobic metabolism and behavior during predatory encounters between snakes ( Thamnophis elgans) and salamanders (Pkthodonjordani) . Oecologia 53: 93-97. Fraadm ETB, 1934. The anatomy o the salamander. London: Oxford University Press. f Cans C. 1960. Studies on amphisbaenids (Amphisbaenia, Reptilia) 1. A taxonomic revision of the f Trogonophinae, and a functional interpretation of the amphisbaenid adaptive pattern. Bulletin o the f American Museum o Natural History 119: 129-204. Canm C. 1973. Uropeltid snakes-survivors in a changing world. Endeavour 32: 60-65. C h a m C. 1974. Biomechanics, an approach to vertebrate biology. Philadelphia: JB Lippincott Co. Hnirmton NC. 1949. The local distribution and ecology of the plethodontid salamanders of the southern Appalachians. Ecological Monographs 19: 47-73. Hatwolc H 1960. Burrowing ability and behavioral responses to desiccation of the salamander, Plethondon . cinereus. Ecology 41: 661-668. SALAMANDER FORM AND FUNCTION 161 Hccht MK, Edwards JL.1977. The methodology of phylogenetic inference above the species level. In: Hecht MK, Goody P, Hecht BM, eds. Major pafferns in verfebrate evolufion. New York: Plenum Press, 3-51. Hildebrand M. 1985. Digging of quadrupeds. In: Hildebrand M, Bramble DM, Liem KF, Wake DB, eds. Funcfional vertebrate morphology. Cambridge, Mass.: Harvard University Press, 89-109. Hinderstein B. 1971. The desmognathine jaw mechanism (Amphibia: Caudata: Plethodontidae). Herpefologica 27: 467-476. Huheey JE. 1964. Use of burrows by the black-bellied salamander. Journal ofthe Ohio Herpetological Socieo 4: 105. Larsen JH Jr, Beneski JT, Jr. 1988. Quantitative analysis of feeding kinematics in dusky salamanders (Desmognathus). (.anadian Journal of (oology 66: 1309- 1 3 I 7. Larsen JH Jr, Cuthrie DJ. 1975. The feeding system of terrestrial tiger salamanders (Ambysfoma figrinum melanosfictum Baird). Journal of Morphology 147: 137- 154. Larsen JH Jr, Bcneski JT Jr, Wake DB. 1989. Hyolingual feeding systems of the Plethodontidae: comparative kinematics of prey rapture by salamanders with free and attached tongues. Journal of Experimenfal (oology 252: 25-33. Larson A. 1991. A molecular perspective on the evolutionary relationships of the salamander families. In: Herht MK, Wallace B, MacIntyre RJ, eds. Evolufionary biology, Vol. 25. New York: Plenum Press, 21 1-277. Lombard RE, Wake DB. 1977. Tongue evolution in the lungless salamanders, family Pethodontidae. 11. Function and evolutionary diversity. Journal tj Morphology 153: 39-80. Lombard RE, Wake DB. 1986. Tongue evolution in the lungless salamanders, family Plethontidae IV. Phylogeny of plethodontid salamanders and the evolution of feeding dynamics. Sysfematic (oology 35: 532-55 I . Martof BS. 1962. Somr aspects of the life history and ecology of the salamander Leurognathus. American Midland Naturalist 67: 1-35. Mnrtof BS, Scott DC. 1957. Ihe food of the salamander Leurognafhus. Ecology 38: 494-501. Miller BT, Larsen JH Jr. 1990. Comparative kinematics of terrestrial prey capture in salamanders and newts (Amphibia: Urodela: Salamandridae). Journal of Experimenfal <oology 256: 135-153. Noble CK. 1927. l h e plethodontid salamanders: some aspects of their evolution. American Museum Novitates, NO. 249: 1-26. Noble GK. 1931. The biology o f f h e Amphibia. New York: The McGraw-Hill Book Co. Nussbaum RA. 1976. Geographic variation and systematics of salamanders of the genus Dicamptodon Strauch (Ambystomatidae). Miscellaneous Publications of fhe Uniuersiy of Michigan, No. 149: 1-94. Olson EC. 1961. Jaw mechanisms: rhipidistians, amphibians, reptiles. American <oologisf 1: 205-21 5. P h t t J. 1935. A comparative study of the hyobranchial apparatus and throat musculature in the Plethodontidae. Journal of Morphology, 57: 21 3-25 I . Pope CH. 1924. Notes on North Carolina salamanders with especial reference to the egg-laying habits of Leurognafhus and Desmognafhus. American Museum Novitafes, No. 153: 1- 15. Pope CH. 1928. Some plethodontid salamanders from North Carolina and Kentucky with the description of a new rare of Leurognafhus. American Museum Novitafes, No. 306: 1-19. Pope CH, Hairston NG. 1947. The distribution of Leurognafhus. A southern Appalachian genus of salamanders. Fieldiana, ~ o o l o g y 31: 155- 162. Regal PJ. 1966. Feeding specializations and the classification of terrestrial salamanders. Evolufion 2& 392-407. Reilly SM,Lauder CV. 1989. Kinetics of tongue projection in Ambysfoma figrinum: quantitative kinematics, muscle function, and evolutionary hypotheses. Journal of Morphology 199: 223-243. Reilly SM,Lauder GV. 1990. The evolution of tetrapod feeding behavior: kinematic homologies in prey transport. Evolution 4: 1542-1557. Roth C. 1976. Experimental analysis of the prey catching behavior of Hydromantes italicus Dunn (Amphibia, Plethodontidae). Journal of Comparative Physiology A 109: 47-58. Schwenk K, Throdunorton CS. 1989. Functional and evolutionary morphology of lingual ferding in squamate reptiles: phylogenetics and kinematics. Journal of <oology, London 219: 153-1 75. Schwenk K, Wake DB. 1988. Medium-independent feeding in a plethodontid salamander: tongue projection and prey capture under water. American <oologisf 28: 115A. Smith HM,Taylor EH. 1948. An annotated checklist and key to the Amphibia of Mexico. Eullcfin o f t h e Unifed States National Museum, No. 194: 1-118. Soler E . 1950. O n the status of the family Desmognathidae (Amphibia, Caudata). Universib of Kansnr Science I Bulletin 33: 459-480. Stebbinn RC. 1951. Amphibians of Western North America. Berkeley: University of California Press. Valentine BD. 1963. The plethodontid salamander Phaeognathus: collecting techniques and habits. Journal of fhe Ohio Herpetological Sociey 4 49-54. 1 Wainwright SA, Biggn WD,Currey JD, G o n h e JM. 1976. Mechanical design in organisms. London: Edward Arnold. Wake DB. 1963. Comparative osteology of the plethodontid salamander genus Aneides. Journal of Morphology 113: 77-1 18. Wake DB. 1966. Comparative osteology and evolution of the lungless salamanders, family Plethodontidae. Memoires of fhe Southern Calijornia Academy of Sciences 4: 1-1 I I . I62 K. SCHWENK AND D. B. WAKE Wake MH. 1993. The skull as a locomotor organ. In: Hanken JH, Hall B, eds. The verfebrate skull, Vol. 25. Funclional and cvolufionary mechanisms. Chicago: University of Chicago Press, 197-240. Wake MH, IknLen J. 1982. Development of the skull of Dcrmophis mexicanus (Amphibia:Gymnophiona), with comments on skull kinesis and amphibian relationships. journal qf Morphology 173: 203-223. Worthington RD, Wake DB. 1972. Patterns of regional variation in the vertebral column of terrestrial salamanders. journal o Morphology 137: 257-278. f
Find millions of documents here - Study Guides, Homework Solutions, Papers, Exam Answer Keys and more.
Course Hero has millions of course related materials that will enable you to learn better, faster and get an A in all your courses.
Below is a small sample set of documents:
SchwenkREVIEWPiankaQRB95.pdf
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 349 Fall, 2008
Path: UConn >> EEB >> 369 Spring, 2008
Path: UConn >> EEB >> 458 Fall, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 155q Spring, 2008
Path: UConn >> PHYS >> 257 Spring, 2008
Path: UConn >> PHYS >> 257 Spring, 2008
Path: UConn >> ENGL >> 1011 Fall, 2008
Path: UConn >> ENGL >> 1011 Fall, 2008
Path: UConn >> ENGL >> 1011 Fall, 2008
Path: UConn >> ENGL >> 1011 Fall, 2008