Dentition and diet in snakes: adaptations to oophagy in the Australian elapid genus Sirnoselups

J. Zool., Lond. (1988) 216, 519-528 Dentition and diet in snakes: adaptations to oophagy in the Australian elapid genus Sirnoselups J. D. SCANLON AND R. SHINE Zoology A08, The University of Sydney, NS W 2006, Australia (Accepted 23 February 1988) (With 2 figures in the text) Most small fossorial proteroglyphous Australian snakes of the genus Sirnoselups feed on adult lizards, but the species of one lineage (the sernfusciutus group) feed exclusively on the eggs of squamate reptiles. Examination of cleared, alizarin preparations showed that dentition of the saurophagous species is similar to that of other elapids, but dentition of the oophagous taxa is highly modified. The anterior (palatine and maxillary) teeth other than the fangs are reduced in size and number whereas those of the pterygoid (and in S. roper?, the dentary) are enlarged posteriorly, becoming compressed along a longitudinal plane and angled medially. The shape of the pterygoid and quadrate is also modified. Two Sirnoselaps species with broader diets (eating both adult lizards and their eggs) show typical saurophagous dentition in one case, oophagous dentition in the other, showing that either type of dentition can be used to capture and ingest either type of prey. We suggest functional explanations for the dentitional modifications in the egg-eating snakes, primarily in terms of the advantages of applying considerable force to the eggshell. Oophagous modifications within Sirnoselups are convergent with those seen in several independently-derived lineages of oophagous colubrid snakes, but (perhaps because of the presence of the fang) differ in having the enlarged blade-like teeth on the pterygoid or dentary rather than the maxilla. Contents Page Introduction... 519 Materials and methods... 520 Results... 521 Discussion... 525 References... 527 Introduction Many species of snakes feed on only one or a few types of prey, and show corresponding modifications of cranial anatomy and dentition. Distantly related taxa of snakes often show remarkable convergences in trophic morphology (e.g. Savitsky, 1983). The best-studied example of this phenomenon is the independent evolution of enlarged venom-conducting fangs in several lineages of snakes (e.g. McDowell, 1986), but even more detailed resemblances have evolved among snakes with highly specialized diets. Perhaps the best example occurs among snakes which feed mainly or exclusively on reptilian eggs. Reduction in size and number of the anterior Present address: Department of Zoology, University of New South Wales, P.O. Box 1, Kensington 519 0952-8369/88/011519+ 10 80340 e) 1988 The Zoological Society of London

520 J. D. SCANLON AND R. SHINE maxillary teeth, accompanied by development of a large blade-like tooth at the rear of the upper jaw, is seen in several oophagous colubrid taxa: for example, Cemophoru of North America, Oligodon of Asia, Prosymna of southern Africa, and Stegonotus of tropical Australasia (Wall, 1923; Palmer & Tregembo, 1970; McDowell, 1972; Broadley, 1979). Current phylogenetic hypotheses (e.g. Cadle, 1987; McDowell, 1986) suggest that these taxa are only distantly related, and hence that these similarities are due to convergent evolution (adaptation to oophagy?) rather than to common inheritance. Recent studies on small fossorial proteroglyphous (front-fanged) snakes of arid and semi-arid regions of Australia have revealed another example of oophagy and associated dental modification (Shine, 1984). Although most species of Sirnoselups feed primarily on lizards, one lineage within the genus (the semifasciatus group) feeds mainly or exclusively on reptilian eggs. Cursory examination revealed enlarged teeth at the rear of the mouth in S. semifasciatus, but not in a saurophagous congener (Shine, 1984). This case is particularly interesting in that it seems to be the only record of specialist oophagy in venomous terrestrial snakes, and involves only one part of a genus, facilitating comparisons of dentition between saurophagous and oophagous congeners. Information on phylogenetic affinities of these small snakes is needed if one is to interpret the evolutionary changes in dentition. Chromosomal, morphological and electrophoretic data ally Sirnoselaps with another small fossorial group (Neelaps), and less closely with an assemblage of other small oviparous Australian elapids (Furinu, Glyphodon and possibly Vermicella) (Mengden, 1985; Wallach, 1985; Scanlon, 1985). The oophagous S. semifasciatus group form a monophyletic lineage derived within the primitively saurophagous genus (Scanlon, 1985). Cladistic analysis of cranial features and external morphology suggests that Simoselaps divides basally into two monophyletic lineages regarded here as subgenera: Simoselaps for the bertholdi group (as in Storr, 1985), and Brachyurophis for all the remaining species (i.e. subgenus Brachyurophis of Storr with the addition of S. warro). Simoselaps warro is similar phenetically to Furina (Storr, 1981), but also shares derived characters with Neelaps, the S. bertholdi group and the clade containing both S. fasciolatus and the S. semifasciatus group. Among the latter group, S. australis is the sister taxon to the other, more derived species (Scanlon, 1985). Materials and methods Observations of skulls, jaws and teeth were based on specimens in the collections of the Australian Museum (Sydney) and the Western Australian Museum (Perth). Specimens were cleared with KOH and trypsin and stained with alizarin, examined under a binocular microscope and drawn using a camera lucida attachment. One to 3 specimens of 9 species of Sirnoselaps were treated in this way, as were 2 V. annulata and 1 specimen each of F. diadema and F. barnardi. Tooth counts were obtained on an additional 6 specimens (2 more species; not cleared and stained) by scraping away soft tissue from tooth-bearing bones under the binocular microscope (specimens listed in Appendix). Nomenclature is a problem with one species, S. roperi from northern WA and scattered localities in the NT, which Storr (1985) treated as synonymous with S. roperi of northern NT (type locality Roper River Mission). The forms, distinguished mainly by the number of bands on the body and tail, have been found in sympatry at several localities (Wells & Wellington, 1985; C. James, pers. comm.; Australian Museum records) and are probably specifically distinct, but no species name is available for the narrow-banded form (Wells & Wellington only succeeded in erecting a junior synonym of S. roperi). Here this species is referred to as S. roper?.

SNAKE DENTITION TABLE I Tooth counts (including unoccupied positions) for specimens of Sirnoselaps and related genera (left sidejright side) 52 1 Species S. bertholdi S. littoralis S. anomalus S. warro S. fasciolatus S. australis S. semifasciatus S. approximans S. roper? S. incinctus S. woodjonesi Neelaps calonotus N. bimaculatus Vermicella annulata Furina diadema F. barnardi Maxilla Palatine Pterygoid Dentary 13/13 11/10 11/12 819 14/14 15/14 14/14 /14 11111 11/11 lojl0 10111 919 lojl0 12/13 15/14 10111 13/12 515 515 817 617 61 15/15 15/19 Results The 11 species of Simoselaps examined seem to form a graded series between relatively generalized small elapids (saurophagous, with only minor skull modifications for burrowing) through to highly specialized oophagous fossorial taxa. The more generalized species are S. warro and S. fasciolatus (subgenus Brachyurophis in part), and the S. bertholdi group (subgenus Sirnoselaps). Dentition in these snakes is fairly similar to that in the related saurophagous genera Neelaps, Furina and Glyphodon (see Shine, 1981, 1984 for dietary data), and also to that of Vermicella, which feeds exclusively on typhlopid snakes (Shine, 1980). In all of these taxa, teeth are retained on the maxilla, dentary, palatine and pterygoid (numbers of teeth are given in Table I). As in elapids generally, there are two canaliculate fangs anteriorly on the maxilla (only one of which is normally ankylosed on each side at any time), followed in most species by a series of small solid teeth similar to the palatine teeth in size and simple shape. In Furina and Neelaps, the fang sockets are nearly transversely aligned, allowing the tips to occur at approximately the same position when either fang is ankylosed (as described for cobras by Bogert, 1943). This is also the case in Simoselaps warro and S. fasciolatus, but in the S. semijasciatus group the fang bases are more obliquely aligned, and this tends to more longitudinal alignment through a series S. australis-

522 J. D. SCANLON AND R. SHINE semifasciatus-upproximans- roper? (Fig. 1). Vermicella also has somewhat obliquely aligned fangs. The diastema is usually a little shorter than the orbit, and considerably longer than the fangs in the fossorial genera (approximately equal to fang length in Furinu, as in typical terrestrial elapids; compare figures in Bogert, 1943; Worrell, 1963 and McDowell, 1967). In each of the palatomaxillary tooth rows, the teeth decrease in size posteriorly, except in oophagous species as described below. Reduction in tooth number, a common trend in fossorial and diet-specialized snakes, has occurred in parallel in Neelups and Simoselaps, so that stages of tooth reduction are not diagnostic of any natural group in this radiation: Furinu, Simoselaps warro and S.Jlzsciolatus have four or more solid maxillary teeth and the palatine extends anteriorly to the fangs, while in the other species both anterior tooth rows are reduced so that there are no more than two solid maxillaries and the palatine lies behind the front of the maxilla. The pterygoid tooth row is also reduced in Vermicellu and Neelups bimuculutus, ending anterior to the ectopterygoid articulation. This reflects the typical condition in fossorial snakes with unspecialized posterior teeth. Neelaps culonotus and the S. bertholdi group also have unspecialized pterygoid teeth, but retain a primitively long tooth row. The dentary tooth row of all of the taxa considered in the present paper consists of two regions, in each of which the teeth first increase, then decrease in size. The two regions are not separated by a gap, but the reduced tooth size in the central region produces a false diastema at approximately the position of the maxillary fang. The anterior dentary teeth in Glyphodon are notably large and erect (so much so that they are responsible for the generic name: glyph-odon = fang-tooth), and a similar condition is seen in Furinu and some Simoselups (e.g. S. fasciolutus: Fig. 2). The first few anterior teeth on the dentary are usually longer and more slender than the posterior ones, but the difference is small in S. semifusciutus (Fig. 2), as in Neelups and the S. bertholdi group. A similar but independent differentiation within the dentition on the dentary occurs in Vermicellu annulatu, where the anterior teeth are much more robust (but not much longer) than the posterior ones; this is similar to the distinction between palatine and pterygoid teeth in this genus. The fossorial, oophagous S. semifasciatus group (australis, semfasciutus, approximans, roperi, nominate roperi of northern NT, incinctus, cumpbelli, woodjonesi) depart considerably from this pattern. All have skull modifications readily interpretable as adaptations for burrowing. These are evident externally and include the large rostral, with a sharp horizontal anterior edge, posterior extension between the internasals, flattened or hollowed ventral surface, and short, conical form of the whole head (Storr, 1967). In the skull, these same features are seen as the enlarged and robust premaxilla stabilized by broad contact with the septomaxillae, broadened nasal-frontal approximation or contact, and relatively short, broad and deep parietal. The anterior maxillary teeth behind the fang are reduced in size and number, whereas those of the pterygoid enlarge posteriorly (the reverse of the condition in other Simoselaps and most snakes), and are compressed along a longitudinal plane and angled medially. This saw-like tooth row continues to the posterior extremity of the pterygoid, which is square posteriorly because the posterolateral angle of the pterygoid flange, which approaches the quadrate as in all snakes, is lateral rather than posterior to the last tooth. The quadrates are short and effectively vertical, not extending beyond the skull (exoccipitals) posteriorly, and the entire braincase is shorter than in the saurophagous Simoselups. Sirnoselups fasciolatus is intermediate in this respect between the S. semfusciatus group and the S. bertholdi group, but retains the generalized dentition. Hence, the tooth row in the oophagous Sirnoselups is very long relative to the skull and quadrates (and pterygoid flange) because of the combined effects of a relative shortening of the cranium and a lengthening of the tooth-bearing area (Figs 1 and 2).

SNAKE DENTITION d 523 /

524 J. D. SCANLON AND R. SHINE m d uj c W n B

SNAKE DENTITION 525 In S. australis and S. semijiasciatus, in many respects the most primitive members of the oophagous lineage (Scanlon, 1985), the pterygoid teeth are relatively few (9 to 11 : Table I) and tend to form a single gradient in size, increasing posteriorly. However, the last few teeth are less bladelike (have a shorter base) and are more recurved. Sirnosetaps fasciolatus also has 11 pterygoid teeth, but these are of the usual conical form and increase in size anteriorly. Furina has 15 or more pterygoid teeth, whereas S. warro has 14 to 15, and S. approximans and S. roperi have 12 to 15. The pterygoid is more strongly modified in the remainder of the oophagous species. In S. approximans the pterygoid row is longer than in S. australis or S. semifasciatus (12 to 13 teeth) and less straight. The difference in tooth form along the pterygoid is more obvious: from small subconical anterior teeth to longer compressed and oblique posterior teeth. None the less, the pterygoid row appears to form a single series with approximately alternate replacement along its length. To an even greater degree than in S. semifasciatus, the last few posterior teeth are smaller, slightly less compressed, and directed more posteriorly. The most highly derived oophagous dentition (of species cleared and stained in this study) is seen in S. roper?. The ptergyoid bears many teeth (14 or 15), with the posterior ones having a blade-like form. However, these teeth are smaller, shorter-based, and more uniform in size than are those of S. approximans. In addition, the dentary teeth are larger (less reduced) than in S. semifasciatus, and are of two different types. The teeth anterior to the false diastema are conical, curved, and in a forwardly erect position. In contrast, the posterior teeth, although recurved and of similar height to the anterior elements, are considerably longer at the base. Hence, their size and laterally compressed shape make them similar to those on the posterior pterygoid. Discussion Although more data on both diets and dentition in Simoselaps would be valuable, the information presented above suggests a strong correlation between food habits and tooth morphology. The saurophagous members of this genus, like related genera with similar diets, have an orthodox elapid dentition. The oophagous lineage within Simoselaps displays a reduction in the number and size of maxillary teeth, and has developed large blade-like teeth on the pterygoid (and in S. roper?, the dentary). The two species of Simoselaps reported to feed on both adult lizards and their eggs (Shine, 1984) are of particular interest. One (S. fasciolatus) has the usual saurophagous dentition, whereas the other (S. australis) has the derived oophagous condition. Cladistic analysis of morphological characters suggests that these taxa may represent degrees of divergence from the ancestral (saurophagous) group leading to the present-day oophagous species (Scanlon, 1985). The ability of snakes with either type of dentition to eat either type of prey is not surprising. Oophagy occasionally occurs among many small elapids with unmodified dentition (Shine, 1984), and the converse is also known: snakes with oophagous dentition consume a wide variety of prey types (e.g. Stegonotus: McDowell, 1972). Presumably, neither type of dentition precludes consuming the wrong prey type, but simply increases efficiency with the correct prey type. How might such advantages and disadvantages work? Evolution of the oophagous dentition involves a reduction of the size and number of the teeth in the front of the mouth, and so is likely to reduce a snake s ability to maintain a firm grasp of squirming scincid prey. On the other hand, the oophagous modifications can be readily interpreted in terms of mechanisms to puncture a tough but flexible eggshell. Swan (1983) described each of two eggshells removed from the stomach of a S. semijiasciatus as having been punctured once with a neat round hole. Presumably, the large triangular teeth are used for this purpose. A large hole is needed in order for the contents of the egg

526 J. D. SCANLON AND R. SHINE to flow out (the shells are not digested: Shine, pers. obs.), and considerable pressure would be needed to make such a hole. We hypothesize that this may explain: (i) the triangular shape of the teeth (resist breaking when force is applied, and make a large incision); (ii) the position of the teeth at the rear of the mouth and with medial alignment (greater mechanical advantage); and (iii) reduction of the size and number of other teeth (so that all of the pressure on the eggshell is exerted over a restricted area, i.e. the edges and tips of the specialized posterior teeth). Observations of captive S. australis feeding on scincid eggs (M. Fitzgerald, pers. comm. with photographs) confirm that the snake brings a great deal of pressure to bear on the egg in order to make an incision. The egg is held in the rear of the mouth while the snake makes active chewing motions until the eggshell is punctured. Further chewing then causes the collapse of the shell, and fluid from the egg can be seen leaking from the snake s mouth. Completion of the swallowing process is then rapid, whereas the initial phase (positioning and puncturing the egg) is more prolonged. Swallowing in snakes is generally accomplished by protraction and retraction of both upper and lower jaw elements of each side alternating with parallel movements of the elements of the other side (Gans, 1961; Cundall, 1987). The operation of saw-like posterior teeth, whether maxillary, pterygoid or dentary, in perforating leathery egg-shells, would not seem to require any modifications to this basic mechanism, except that the prey item (egg) is not swallowed directly but retained in the mouth while jaw protractions and retractions continue until the enlarged teeth puncture or tear the shell sufficiently for injection of the contents. Behaviourally, this phase of holding and active chewing (as observed in S. australis) is similar to the post-strike holding or envenomation phase (usually, up to the beginning of prey orientation) in typical saurophagous elapids, but it occurs after orientation and after the commencement of swallowing. Neither is there a strike as such, but this also is often lacking in saurophagous snakes taking dead or inactive prey, including squamate eggs. The loose connection of palatine and pterygoid in the S. semifasciatus group probably allows increased flexure of the palatine below the level of the articulation, and lack of teeth on the posterior half of the palatine suggest that the separated tooth rows may be able to engage the prey surface at different angles and thus accommodate the curvature of an item such as an egg. However, this feature is not necessarily related to oophagy since a very similar condition of the palatine is seen in Neelaps himaculatus which feeds only on elongate lizards (Shine, 1984). The oblique or longitudinal alignment of the fang bases in the oophagous group represents a reversion toward the lizard or primitive-snake condition in which the maxillary teeth form an undifferentiated and continuous series, but not to the extent seen in Toxicocalamus and Ogmodon (McDowell, 1969). Assuming that inward flexure of the anterior maxilla is of value in maintaining uniformity and efficiency of the venom-delivery system, it seems that a dietary shift from active prey to eggs-not requiring effective envenomation-might free the maxilla from such selective pressure. Instead, selection for increased gape might result in straightening of the anterior maxilla, i.e. lateral displacement of the enlarged first tooth (which still retains the structure, but apparently not the function, of a fang). Again, doubt is cast upon this interpretation by a similar change evident in a (probable) relative, Vermicella (Scanlon, 1985), which feeds on typhlopids (Shine, 1980) and presumably retains envenomation as a component of its feeding behaviour. Feeding by Vermicella has not been observed, however, and it is conceivable that feeding normally takes place in tunnels that simplify prey capture and orientation such that envenomation is not critical. The dentitional similarity between the oophagous Simoselaps and various oophagous genera of colubrid snakes (see Introduction) is impressive. All possess large blade-like teeth at the rear of the

SNAKE DENTITION 527 mouth, although in each of the colubrids the modified teeth are on the posterior maxilla. In Sirnoselaps, the enlarged teeth are on the pterygoid or dentary. It may be that the presence of the fang on the anterior maxilla is in some way related to this difference. Unfortunately, we cannot even determine whether reduction of the maxillary teeth occurred before or after the origin of compressed pterygoid teeth (or if they occurred simultaneously), because no intermediate condition between that of S.,fasciolutus and S. australis is known. We thank G. M. Storr and A. E. Greer for permission to borrow and prepare specimens, M. Fitzgerald for a careful and well-documented description of feeding by his captive snakes and D. T. Anderson and M. Stevens for the loan of microscope equipment. REFERENCES Bogert, C. M. (1943). Dentitional phenomena in cobras and other ekdpids with notes on adaptive modifications of fangs. Bull. Am. Mus. nut. Hist. 81: 285-360. Broadley, D. G. (1979). Predation on reptile eggs by African snakes of the genus Prosymna. Herpetologica 35: 338-341. Cadlc, J. E. (1987). Geographic distribution in phylogeny and zoogeography. In Snakes; ecology and evolutionary biology: 77-105. Seigel, R. A., Collins, J. T. & Novak, S. S. (Eds). New York: Macmillan. Cundall, D. (1987). Functional morphology. In Snakes: ecology and evolutionary biology: 106-140. New York: Macmillan Publ. Co. Gans, C. (1961). The feeding mechanism of snakes and its possible evolution. Am. Zoo/. 1: 217-227. McDowell, S. B. (1967). Aspidomorphus, a genus of New Guinea snakes of the Family Elapidae, with notes on related genera. J. Zool., Lond. 151: 497-543. McDowell, S. B. (1969). Toxicocalamus, a New Guinea genus of snakes of the family Elapidae. J. Zool., Lond. 159: 443-511. McDowell, S. B. (1972). The species of Stegonotus (Serpentes, Colubridae) in Papua, New Guinea. Zool. Meded.. Leiden 47: 6-26. McDowell, S. B. (1986). The architecture of the corner of the mouth of colubroid snakes. J. Herpet. 20: 353-407. Mengden, G. A. (1985). Australian elapid phylogeny: a summary of the chromosomal and electrophoretic data. In Biology of Australasian frogs and reptiles: 185-192. Grigg, G., Shine, R. & Ehmann, H. (Eds). Roy. Zool. SOC. N.S.W., Sydney. Palmer, W. M. & Tregernbo, G. (1970). Notes on the natural history of the scarlet snake Cemophora coccinea copei Jan in North Carolina. Herpetologica 26 300-302. Savitzky, A. H. (1983). Coadapted character complexes among snakes: fossoriality, piscivory, and durophagy. Am. Zool. 23: 397-409. Scanlon, J. D. (I 985). Phylogeny and relationships of the elapid snake genus Simoselaps Jan, 1859: the evolution of a group of burrowing snakes. Unpubl. Hons thesis, School of Biological Sciences, University of Sydney. Shine, R. (1980). Reproduction, feeding and growth in the Australian burrowing snake Vermicella annulata. J. Herpet. 14 71-77. Shine, R. (1981). Ecology of Australian elapid snakes of the genera Furina and Glyphodon. J. Herpet. 15: 219-224. Shine, R. (1984). Ecology of small, fossorial Australian snakes of the genera Neelaps and Simoselaps (Serpentes: Elapidae). In Vertebrate ecoiogyundsystematics-a tribute tohenry S. Firch: 173-183. Seigel, R. A,, Hunt, L. E., Knight, J. L., Malaret, L. & Zuschlag, N. L. (Eds). Mus. Nat. Hist., Univ. of Kansas, Lawrence. Storr, G. M. (1967). The genus Vermicella (Serpentes: Elapidae) in Western Australia and the Northern Territory. J. Proc. R. Soc. West. Aust. 5080-92. Storr, G. M. (1981). The genus Furina (Serpentes: Elapidae) in Western Australia. Rec. W. Aust. Mus. 9 119-123. Storr, G. M. (1985). Phylogenetic relationships of Australian elapid snakes: external morphology with an emphasis on species in Western Australia. In Biology of Australasian frogs andreptiles: 221-222. Grigg, G., Shine, R. & Ehmann, H. (Eds). Roy. Zool. SOC. N.S.W., Sydney. Swan, M. (1983). Notes on the half-girdled snake Simoselaps semfusciatus (Gunther). Herpetofauna 1494. Wall, F. (1923). A review of the Indian species of the genus Oligudon, suppressing the genus Simotes (Ophidia). Rec. Ind. Mus. 25: 305-334.

528 J. D. SCANLON AND R. SHINE Wallach, V. (1985). A cladistic analysis of the terrestrial Australian Elapidae. In Biology ofaustralasianfrogs and reptiles: 223-253. Grigg, G., Shine, R. & Ehmann, H. (Eds). Roy. 2001. SOC. N.S.W., Sydney. Wells, R. W. & Wellington, C. R. (1985). A classification of the Amphibia and Reptilia of Australia. Aust. J. Herpet. (Suppl.) 1: 1-98. Worrell, E. (1963). Reptiles of Australia. Sydney: Angus & Robertson. Appendix List of specimens examined Species n Registration (a) Cleared and stained specimens Simoselaps bertholdi 2 S. littoralis 2 S. anomalus 1 S. warro 2 S. fasciolatus 2 S. australis 1 S. semifasciatus 3 S. approximans 1 S. roperi 1 Neelaps calonotus 2 N. bimaculatus 2 Furina diadema 1 F. barnardi 1 Vermicella annulata 2 WAM R 20578, R 28174 WAM R 86866-7 WAM R 13815 AM R 14395, R 19017 WAM R 5936,R 8389 AM R 98167 WAM R 22882, R 59218; AM No Data WAM R73478 WAM R 20349 WAM R 592, R 5816 WAM R 20604, R 40978 AM R 98165 AM No Data AM R 21345. AM No Data (b) Specimens with tooth counts only S. warro AM R 46024 S. approximans AM R 71624 S. roperi AM R 70025 S. inrinctus AM R 64014 S. woodjonesi AM R 30337 V. annulata Shine teaching specimen, No Data

本文献由 学霸图书馆 - 文献云下载 收集自网络, 仅供学习交流使用 学霸图书馆 (www.xuebalib.com) 是一个 整合众多图书馆数据库资源, 提供一站式文献检索和下载服务 的 24 小时在线不限 IP 图书馆 图书馆致力于便利 促进学习与科研, 提供最强文献下载服务 图书馆导航 : 图书馆首页文献云下载图书馆入口外文数据库大全疑难文献辅助工具