Shell microstructure and morphogenesis of the ornamentation in Cymatoceras Hyatt, 1883, Cretaceous Nautilida. Systematic implications RÉGIS CHIRAT AND HUGO BUCHER LETHAIA Chirat, R. & Bucher, H. 2006 03 31. Shell microstructure and morphogenesis of the ornamentation in Cymatoceras Hyatt, 1883, Cretaceous Nautilida. Systematic implications. Lethaia , Vol. 39, pp. 57
/64. Oslo. ISSN 0024
/1164. The transverse ridges or so-called ‘ribs’ of Cretaceous Cymatoceras correspond to the juxtaposition of thick, projected and imbricated radial tile-shaped lamellae of outer prismatic layer. Each of these lamellae formed in a cycle including an outward extension and secreting phase of the mantle edge, followed by his temporary withdrawal behind the periphery of the lamella. The sculpture of the underlying nacreous layer reflects only the underlying morphology of the juxtaposed adapical parts of the imbricated lamellae of outer prismatic layer. On internal moulds, the subdued symmetric undulations, which reflect the internal surface of the nacreous layer, can easily be mistaken for imprints of comarginal ribs (this last term is restricted here to undulations of the outer shell without structural discontinuities other than growth lines). Distinction of this kind of ornamentation challenges the monophyly of the family Cymatoceratidae, classically interpreted to include all ‘ribbed’ post-Triassic Nautilida, the so-called ‘ribbing pattern’ encompassing lamellae, fasciculate growth lines and divaricate ornamentation. Whether or not radial tile-shaped lamellae of outer prismatic layers a synapomorphy of an emended ‘cymatoceratid’ clade cannot be solved until its history can be traced through a well-corroborated phylogeny, allowing in turn the evaluation of the hypothesized heterochronic (paedomorphic) shifting of the embryonic growth pattern trough postembryonic development. I Cephalopoda, microstructure, morphogenesis, Nautiloids, systematics. Régis Chirat [chirat@univ-lyon1.fr], CNRS UMR 5125, Universite´ Claude Bernard Lyon 1, 2 rue Dubois, F-69622 Villeurbanne Cedex, France. Hugo Bucher [Hugo. FR.Bucher@pim. unizh.ch], Paläontologisches Institut und Museum, Universität Zürich, Karl Schmid-Strasse 4, CH-8006 Zürich, Switzerland; 1st November 2004, revised 25th October 2005. Over the past two decades, it has become patent that soundness of evolutionary process-related hypothesis depends upon well-corroborated phylogeny. No matter how a phylogenetic analysis is conducted, it is preceded by a character definition in a way that identifies the ‘same’ traits in different taxa. Formalizing this hypothesis of primary homology is not an unequivocal matter, and depends on some profound issues such as character conceptualization and definition (see Wagner 2001), not to mention the definition of homology itself, which has proven to be among the most difficult of all concepts in comparative biology (e.g. Abouheif et al . 1997; Wray & Abouheif 1998). The reliability of any phylogenetic analysis is related to the number of heritable, discrete and independent characters employed. These problems are especially critical for extinct Nautilida because we confront hypotheses of relationships among species often displaying only slight continuous morphological differences in their shell. This shell is in turn, a highly developmentally integrated buoyancy device, which poses the difficult problem of character correlation. Development underlies evolutionary changes, and investigating this issue at our own palaeontologic level of explanatory mode appears to be the most promising area of research in the definition of appropriate units of phylogenetic analysis. It may lead to the identification of potentially useful similarities among taxa on the basis of their underlying morphogenetic processes. However, character description is influenced by preservation through taphonomic and diagenetic loss of information. Most Nautilida are preserved as internal moulds and the extent to which preservation may bias morphological analyses, and hence phylogenetic inferences, remains a largely underestimated issue. Cymatoceras provides a case study illustrating some of these questions. As classically defined, Cymatoceras is among the most common and geographically widespread post-Triassic Nautilida. Some 80 species ranging from Upper Jurassic to Oligocene have been assigned to this genus. Hyatt (1883, p. 301) defined Cymatoceras as including ‘‘Cretaceous species of the Radiati, remarkable for their transverse costae’’ and designated Nautilus pseudoelegans DOI 10.1080/00241160600582069 # 2006 Taylor & Francis 58 R. Chirat & H. Bucher d’Orbigny, 1840, from the Lower Cretaceous (Neocomian) of France as type species. Cymatoceras was chosen by Spath (1927) as type genus of his family Cymatoceratidae, interpreted to include all ‘ribbed’ post-Triassic Nautilida. This taxon has remained interpreted essentially as conceived by Spath (Kummel 1956, 1964; Matsumoto & Muramoto 1983; Dzik 1984), while the validity of grouping of all post-Triassic ‘ribbed’ genera into a single family has been questionned by others (Wiedmann 1960; Tintant 1969, 1984, 1989, 1993; Tintant & Kabamba 1983; Bardhan & Halder 2000). Well-preserved specimens of Cymatoceras allow investigation of shell wall microstructure and the morphology of the so-called ‘ribs’. This ornamentation is compared to that of other post-Triassic genera, which provides new insights into systematics of Nautilida. Material Observation of the external morphology of the ornamentation in Cymatoceras is mainly based on three wellpreserved specimens of Cymatoceras varusensis (d’Orbigny, 1850) from the Upper Neocomian of France, two of them being figured. One specimen of Cymatoceras sakalavum Collignon, 1949, from the Albian of Madagascar allows further investigations of the shell microstructure in cross-section. All figured specimens are housed in the palaeontological collections of the Claude Bernard University, Lyon I, France. Shell microstructure and morphogenesis of the ornamentation At a first glance, the transverse ridges of Cymatoceras varusensis (Fig. 1) could be simply described as comarginal ‘ribs’. Closer examination of the outer shell surface reveals that these ornamental features are in fact overlapping shelly extensions, delineating small adorally open cavities. Cross sections in Cymatoceras sakalavum (Fig. 2) show that this unusual morphology results from the imbrication of thick projected and overlapping radial tile-shaped lamellae of the outer prismatic layer (bilayered), with sigmoid transverse outline. Each lamella arises from beneath the preceding one with a sigmoid shape, projects forwards and gently arches adorally. The adoral limit of each lamella is not in contact with the next one. The shell displays this peculiar ornamentation throughout its whole ontogeny, from embryonic to adult stages. In the embryonic shell, closely spaced and thin overlapping radial lamellae of the outer prismatic layer intersected with fine spiral LETHAIA 39 (2006) lirae, result in the formation of a reticulate pattern (Fig. 3). The transversal ornamentation of Cymatoceras will no longer be described as ribs, since the usage of this term among malacologists is general enough to be applied to almost all types of ornamentation. The morphology of these lamellae is clearly different from that of other types of ornamentation such as that seen in the juvenile specimens of the Carixien species, Cenoceras annulare (Phillips). In this species, the ornamentation corresponds to comarginal undulations of the outer shell without major structural discontinuity, which in turn corresponds to one of the most widespread ornamental feature in molluscs. In order to avoid any confusion, the term rib is used only in this last sense throughout this paper, unless it is put in inverted commas. The unusual microstructure of the shell wall in Cymatoceras shows that lamellae have grown by successive outward extension followed by a temporary withdrawal of the mantle edge behind the periphery of the lamella, resulting in the formation of a small overhanging portion of the shell. Although the temporary withdrawal of the outer mantle fold was probably controlled by the mantle musculature, the question remains open as to what, in turn, activated this retraction. The ornamentation of the underlying nacreous layer does not faithfully reflect the inner surface of the outer prismatic layer, but only the image of the juxtaposed adapical parts of each lamella. However, tileshaped lamellae of the outer prismatic layer similar to that of Cymatoceras varusensis and C. sakalavum may be inferred when the nacreous layer alone is preserved, be it recrystallized or not. In molluscs, indeed, comarginal undulations of the shell wall without major structural discontinuity (i.e. ribs) may be viewed as the expression of oscillatory mechanisms, and as such, obey a construction rule: the amplitude is inversely proportional to the frequency, and the inter-ribs space equates with ribs width. It is not the case for the undulations of the nacreous layer of Cymatoceras. These ‘pseudoribs’ are separated by interspaces shorter than their own width. This morphology is both theoretically and morphogenetically possible only if oscillatory mechanisms display repeated episodic interruptions. Thus, successive outward extension and temporary withdrawal of the mantle edge explain why Cymatoceras is so densely ornamented. The repeated episodic interruption of the secretion of the outer prismatic layer and withdrawal of the mantle edge also explains why ‘pseudoribs’ of the nacreous layer display an asymmetric wavy transverse outline, their adoral side being steeper than their adapical side. Their flattened outline reflects the forward projection of the adoral part of each lamella of the outer prismatic layer. LETHAIA 39 (2006) Shell microstructure and morphogenesis of the ornamentation in Cymatoceras A C 59 B D Fig. 1. Cymatoceras varusensis (d’Orbigny, 1850) (Upper Neocomian, Plan de Revel, Alpes-Maritimes, France). I A, B. EM 20327 (A/1; B/2). I C, D. EM 20328 (C /1; D /2). The width of the lamellae is nearly constant from the umbilical shoulder to the venter, so that bifurcations take place on the flanks, and accommodate the difference between the internal and external spiral length. Thus, up to four ventral lamellae may merge into a single one near the umbilical shoulder. The variable position of branching points of lamellae, between midflanks and the umbilical margin (variation also visible on the nacreous layer) is associated with these major shell discontinuities. In ammonoids, branching points between ribs are aligned along a spiral line (as modeled by threshold gradient in the reaction-diffusion simulations of Hammer & Bucher 1999) while variable position of these points results from the presence of major discontinuities in the shell, as documented in some ribbed ammonoids possessing megastriae (Bucher et al . 1996). In short, flattened and asymmetric undulations of the nacreous layer, with a steeper adoral side and separated by short concave interspaces, accompanied by a variable position of branching points, strongly suggest the presence of imbricated tile-shaped lamellae of outer prismatic layer. This distinctive mode of growth can thus be inferred when the nacreous layer alone is preserved, be it recrystallized or not. On internal moulds, however, the outline of these ‘pseudoribs’ is levelled off by the thick nacreous layer, so that they may appear as scarcely impressed symmetric (or nearly so) undulations, which can easily be mistaken for imprints of ribs. The holotype of Nautilus pseudoelegans d’Orbigny, 1840 (see Kummel 1956, pl. 16, figs 1
/2) was available for the present study. Its shell is partly preserved but the outer prismatic layer is not preserved. The ornamentation, however, consists of flattened and asymmetric undulations, separated by concave interspaces shorter than their own width, with branching points at variable heights on flanks, which obviously corresponds to the ‘pseudoribs’ of the nacreous layer, visible when overlapping tile-shaped lamellae of the outer prismatic layer are eroded away or dissolved. Thus, tile-shaped lamellae of the outer prismatic layer can be considered as a diagnostic character of Cymatoceras. 60 LETHAIA 39 (2006) R. Chirat & H. Bucher oy op na im A B Fig. 2. IA. Cymatoceras sakalavum Collignon, 1949 (FSL 349570, Albian, Ambarimaninga, Madagascar). Cross section of the shell wall showing the thick projected and overlapping radial tile-shaped lamellae of the outer prismatic layer (op) and the underlying nacreous layer (na) (im: inner mould; oy: encrusting oyster). Oral side on the left (Scale bar: 10 mm). IB. Polarized light micrograph (composite view) and ouline of a lamellae of outer primatique layer. The boundary between sublayers of the outer primatique layer is in broken line (Scale bar: 5 mm). Ward (1987, 1988) and Ward & Saunders (1997) documented the variation in shell wall microstructure in Nautilus and Allonautilus . In Nautilus macromphalus and N. pompilius , the periostracum is a very thin (1
/ 5 mm thick) continuous sheet resting on the outer prismatic layer. This disposition is the same in N. belauensis even though the periostracum is much thicker (about 1 mm). In Allonautilus scrobiculatus the periostracum is strikingly different, being composed as a series of long sheets (up to 10 mm in length) extending A outward from the shell wall and splitting into numerous thin layers from a thicker, single layer emerging from between increments of the outer prismatic layer. This discontinuous outer prismatic layer shows some resemblance with that of Cymatoceras. In Allonautilus scrobiculatus , however, these discontinuities are fare more closely spaced, and the outer shell does not display thick projected tile-shaped lamellae as in Cymatoceras . The shell surface of Allonautilus scrobiculatus is smooth, and the imbrications can be seen only with SEM in B Fig. 3. I A, B Embryonic stages of Cymatoceras sakalavum Collignon, 1949 (FSL 349570, Albian, Ambarimaninga, Madagascar), showing closely spaced and thin overlapping radial lamellae of the outer prismatic layer intersected with fine spiral lirae . LETHAIA 39 (2006) Shell microstructure and morphogenesis of the ornamentation in Cymatoceras cross-section. Whether or not the periostracum in Cymatoceras extended outward from the outer shell wall in the same way as in Allonautilus still remains an open question. An analogous structure of the outer prismatic layer has been recognized in Calliphyloceras , a Jurassic phylloceratid (Bucher et al . 2003). However, the nacreous layer of Calliphylloceras bears radial lirae instead of ‘pseudoribs’. This difference is readily explained by the different shape of the imbricated lamellae of outer prismatic layer. In Calliphylloceras , the adapical part of lamellae is flat instead of concave, and the junction line between two consecutive lamellae is at the bending point of the lamellae instead of underneath the convex, adoral part as in Cymatoceras . Systematic implications Spath (1927) presented a very short review and one of the first classifications of post-Triassic Nautilida, including five families and 29 genera. Among them, six new genera, Paracymatoceras, Procymatoceras, Eucymatoceras , Anglonautilus , Cymatonautilus and Syrionautilus , are brought together with Cymatoceras in the family Cymatoceratidae. Spath did not give any diagnosis of his new taxa but Cymatoceras was obviously chosen as the name-bearing type of this family interpreted to include all ‘ribbed’ post-Triassic Nautilida. This taxon was lowered at the subfamily rank by Kummel (1956), but remained interpreted essentially as conceived by Spath, except for the inclusion of other ‘ribbed’ post-Triassic forms, namely Heminautilus Spath, 1927 (previously assigned by Spath to his family Paracenoceratidae), and two new genera, Epicymatoceras and Deltocymatoceras , the description and diagnosis of these last two genera relying exclusively on the descriptions and drawings previously published by Binckhorst (1861) and Schlüter (1876). Kummel (1956, p. 352) interpreted Cymatoceratinae as ‘‘a single genetic unit . . . having as the basic common denominator the ribbing pattern’’ and considered the wide range in variation of shell shape and suture line within this subfamily as reflecting ‘‘a broad adaptive radiation which produced numerous homeomorphs of other genera of the Nautilidae’’. One might wonder why Kummel interpreted the presence alone of an ornamentation as an evidence of close phylogenetic relationships among these ornamented genera, since the so-called ‘ribbing pattern’ in fact encompasses ornamental features, described by Kummel himself as ‘ribs’, ‘fold-like undulations’, ‘V-shaped ribs’, or ‘ribs’ in Procymatoceras that ‘‘appear to be fasciculate growth lines and may not be homologous with those of typical Cymatoceras’’ 61 (Kummel 1956, p. 428). Kummel did not further discuss character homology although, in such cases, this morphology objectively challenges the monophyly of the Cymatoceratinae. In fact, Kummel’s argument for nautilid classification was focused on characterizing perceived evolutionary patterns and was grounded in a broad adaptationist reasoning. In his introductory chapter on classification of post-Triassic Nautiloids, Kummel (1956, p. 327) had set out the groundwork of his view of their evolution in assuming that ‘‘distinctive radiations are recognizable, each representing an elaboration of particular morphological characters . . . . The time-space relationships of the post-Triassic nautiloid ‘species’ and their morphological modifications, interpreted in terms of adaptive radiation, make possible a constructive phylogenetic interpretation of the group’’. Although he acknowledged (Kummel 1956, pp. 331, 346) the ‘‘complete lack of understanding of the adaptive values of the various morphological features’’, he assumed that ‘‘it seems only logical to conclude that the various conch shapes, etc., represent adaptations to a specific niche in the marine environment’’ which leads him to suggest that (Kummel 1956, p. 360) ‘‘the subfamily units are adaptive trends interpreted for the most part on the basis of single characters
/ in the Cymatoceratinae it is the presence of ribbing’’. In the Treatise on Invertebrate Paleontology, Kummel (1964) followed the same line of argument and grouped all post-Triassic ‘ribbed’ genera into the family Cymatoceratidae. Although Tintant proposed a strikingly different interpretation, his classification is also dominated by the asserted adaptive significance of characters. Tintant challenged the monophyly of Cymatoceratidae, assuming that transversal ornamentation was convergently acquired among lineages of smooth species as an adaptation to increased shell strength in high-energy shallow environments, negating any close phylogenetic relationships among ‘ribbed’ genera or species. This premise led him to suggest that the ‘taxonomic value’ of ‘ribs’ that distinguish some species from overall similar but smooth species should be more appropriately relegated at the lower taxonomic level of the sub-genus of contemporary smooth genera (Tintant 1969), or could be reliable at the genus or species level, when this character does not relate to intraspecific variations (Tintant 1993). In several places, however, Tintant openly stated that his proposed classification does not depict phylogeny. He did not further address this dilemma and provided trees (Tintant 1969, 1984, 1993) illustrating his conception of the phylogenetic relationships of the ‘ribbed’ genera, some of them (Procymatoceras , Cymatonautilus, Cymatoceras, Paracymatoceras, Anglonautilus ) arising iteratively from contemporaneous smooth genera. Tintant (1989, p. 363) 62 R. Chirat & H. Bucher stated that ‘‘it is simpler, indeed, and more economic to admit a possibility of appearance of such a type of ornamentation in all post-Triassic Nautilaceae, which seems closely related to growth lines corresponding to the periodic (daily?) increase of the shell. One notes, indeed, a very variable relief of these growth lines among species: sometimes almost invisible to the naked eye, they often strengthen to form small and simple ribs, sometimes gathered on prominents folds on the test’’ (our translation). In fact, Tintant mistakenly considered all ornamental features with simple growth rings seen when growth lines merge. This interpretation, however, applies to Procymatoceras. As noted previously, Kummel (1956) also described the ornamentation of this genus as fasciculate growth lines and pointed out that the ‘ribbing’ is most distinct on the living chamber of Procymatoceras. Emergence of similar transversal ornamentation on the body chamber of adult specimens is a common but greatly variable feature among Nautilida and could be correlated with decrease in growth rate. These growth rings correlated with decrease in growth rate are common among other molluscs. In the other genera assigned to the family Cymatoceratidae, the ornamentation is not restricted to the body chamber of adult specimens. Whether or not transversal ornamentation in some of these genera (Paracymatoceras, Epicymatoceras, Deltocymatoceras and Syrionautilus ) is similar to that of Cymatoceras cannot be discussed further, however, from the available bibliographic data. Nevertheless, we may briefly review some empirical evidence of overlapped radial tileshaped lamellae in at least three genera that is Neocymatoceras , Cymatonautilus and Anglonautilus . Neocymatoceras Kobayashi, 1954, was established for a single specimen from the Oligocene of Kyushu, Japan. The ornamentation of the type-species, Neocymatoceras tsukushiense , was described by Kobayashi (1954, p. 19) as ‘‘numerous flat-topped flexuous radial ribs, or better to say, elevated bands which are separated from one another by narrow grooves’’ with ‘‘bifurcation somewhere on the flank between the middle and one-third from the umbilicus to the venter’’. To judge from the figurations, the ornamentation of Neocymatoceras which can be described as flattened and asymmetric undulations, with steeper adoral side and separated by interspaces shorter than their own width may correspond to imprints of tile-shaped lamellae of the outer prismatic layer similar to that of Cymatoceras . Tintant (1969, pp. 30
/32; pl. B, figs 2
/3; pl. D, fig. 4a
/c; pl. E, fig. 3a
/b) described three specimens referred to as a new species, Cymatonautilus collignoni , from the Lower Callovian of Madagascar and described their ornamentation as fine, closely spaced, angular, sometimes lamellate ribs. Tintant (1987, pp. 114
/115; LETHAIA 39 (2006) pl. 11, fig. 2; pl. 12, fig. 2) also described the ornamentation of four specimens of this species from the Middle Callovian of Saudi Arabia as very fine, numerous and of lamellar aspect. These specimens were available for the present study. The shell surface displays thin overlapping radial lamellae of the outer prismatic layer. These lamellae are more closely spaced than in Cymatoceras and nearly strait in transverse outline. Matsumoto et al . (1984, pp. 289
/292, pl. 57
/58) described and figured two well-preserved specimens from the Cenomanian of Hokkaido (Japan) referred to as Anglonautilus japonicus . As far as one can judge from the plates only, the specimen described as paratype 1, with a partially preserved shell, displays flattened and asymmetric undulations, with steeper adoral side and separated by interspaces shorter than their own width, which suggest tile-shaped lamellae of the outer prismatic layer in Anglonautilus similar to that of Cymatoceras. The ornamentation of two other genera (Eucymatoceras and Heminautilus ) can not be confused with the preceding ones. Eucymatoceras displays an unusual divaricate ornamentation described by Kummel (1956, pp. 431
/433) as ‘‘prominent ribs that form a deep angular V-shaped ventral sinus and on the whorl sides a similar but asymmetrical salient’’. In bivalves, repaired injuries result in gaps and unconformities in the spiral extension of this kind of divaricate pattern (Seilacher 1984). Repeated and frequent withdrawals of the mantle margin associated with the secretion of tile-shaped lamellae of the outer prismatic layer similar to that of Cymatoceras preclude the development of this kind of continuous spatial patterning of a spiral extension such as that seen in Eucymatoceras. Among the species quoted by Kummel (1956, p. 435) as an evidence that ‘‘sinuous cymatoceratid-like ribbing is characteristic of Heminautilus’’, are those from the Aptian of Suez referred to as Nautilus lallieri d’Orbigny by Douvillé (1916, pp. 129
/132, pl. 12, figs 2
/6). Four of the five specimens figured by Douvillé were available for the present study, only the one illustrated in Douvillé’s figure 5 not being found. Judging from the figure only, this last specimen does not display any ornamentation. The three smaller specimens display transversal ornamentation, the smallest of them showing that the ornamentation appears immediately after hatching, marked by the nepionic constriction at approximately 20 mm in shell diameter. However these smaller specimens retain only poorly preserved remains of shell, so that the geometric relationship between growth lines and ornamentation cannot be further examined. The shell of the specimen illustrated in figure 6 of Douvillé’s study is very well preserved. Contrary to what Douvillé wrote, and what was subsequently believed, the ornamentation of this specimen is not comarginal, but originates from the ventral area and LETHAIA 39 (2006) Shell microstructure and morphogenesis of the ornamentation in Cymatoceras diverges adorally toward the flanks, with an outline oblique to growth lines. The angle between these divarications and growth lines decline ventro-dorsally. They originate with a value of about 208 relative to growth lines and disappear at the level of the ocular sinus, where they become parallel to shell margin. They are absent in the dorsal part of the flanks. Their asymmetric transverse outline, with an abrupt adoral side and a gentle adapical slope, approximates the socalled ‘paradigm of the cross-country ski’ assumed to be an evidence of their adaptive nature in bivalves and related to burrowing-enhancing function (e.g. Seilacher 1984). From a morphogenetic point of view, the spiral extension of this divaricate ornamentation (some of them extending through an angular length of more than 408) is incompatible with tile-shaped lamellae of the outer prismatic layer such as that seen in Cymatoceras, which is supported by the locally well preserved growth lines of the outermost calcified shell layer in the specimen of Heminautilus . Available microstructural and morphological evidence does not support monophyly of the Cymatoceratinae(-dae) as defined by Kummel (1956, 1964). The projected and overlapped radial tile-shaped lamellae of the outer prismatic layer are morphogenetically distinct from divaricate ornamentation and fasciculate growth lines. There is no longer any reason, however, to support the antagonistic view of Tintant (1969, 1984, 1989, 1993), of ornamented nautilid evolution as strongly iterative. Whether or not tile-shaped lamellae (which cannot be confused with simple growth rings) are a synapomorphy of an emended ‘cymatoceratid’ clade cannot be solved until its history can be traced through a well-corroborated phylogeny. Above all, this issue implies an accurate morphological study of all ornamented genera which could provides, in turn, new insights into the origin of these tile-shaped lamellae. Tintant (1989) proposed an heterochronic hypothesis to account for the origin of transversal ornamentation stating that phylogenetically, it always appears on the body chamber of mature specimens and progresses by acceleration up to younger stages. In contrast, Bardhan & Halder (2000) suggested that ornamentation, described as ‘fasciculate ribs’, could have originated in first place in a Bathonian species referred to as Paracenoceras jumarense as a result of size reduction during paedomorphosis. Although much smaller than its presumed ancestor, P. calloviense , these authors assumed that P. jumarense evolved through neoteny rather than progenesis, hypothesizing arbitrarily that similar numbers of septa in both species may imply similar life span. In order to explain the absence of ornamentation in P. calloviense , they assumed that ‘ribs’ evolved in P. jumarense as a fabricationnal by-product of a slower growth rate and concluded (Bardhan & Halder 2000, 63 p. 162) that ‘‘initially, the ribs were weak and localised, either on the flanks or the venter of the adult body chamber . . . Subsequently, these ribs became progressively stronger, extending all around the shell like those of contemporary ammonites’’. Although these interpretations are antithetical from a heterochronic standpoint (paedomorphosis versus peramorphosis), transversal ornamentation is thought to appear in both cases as a result of the addition of new terminal stage to the ancestral ontogeny. In both cases, these authors mistakenly considered fasciculate growth lines seen on the body chamber of many species and ornamentation of genera such as Cymatoceras as homologues. On the other hand, tile-shaped lamellae of the outer prismatic layer of Cymatonautilus collignoni and Cymatoceras, and even more those of the embryonic stages of this last genus show an obvious resemblance with the ornamentation of the early embryonic shell of a wellpreserved Toarcian Cenoceras ? sp. (see Chirat & Boletzky 2003, fig. 1). The apex of this embryonic shell is capshaped with a scar-like feature at the top, the so-called cicatrix, corresponding to the initial site of shell deposition as demonstrated by Arnold & Carlson (1986) in Nautilus. The cicatrix area is a lentoid elevation without growth lines, surrounded by a slight peripheral constriction. Originating approximately at the margin of this constriction, spiral lirae radiate away from the apex of the shell and are intersected by fine growth lines and wider, regularly spaced transversal lines of ornamentation. Between two neighbouring spiral lirae, the adoral part of the transversal lines of ornamentation dip back, resulting in the formation of small overhanging portion of the shell. In Nautilus , the peripherial constriction surrounding the cicatrix area marks the boundary between two stages of early shell development: an earlier cicatrix-mode of shell secretion by the shell gland and an accretionary growth at the mantle margin, as indicated by the first appearance of growth lines. These two basic modes of shell secretion reflect the difference in the degree of differentiation of the mantle during early embryogenesis (Arnold & Landman 1992; Tanabe & Uchiyama 1997). In the Toarcian embryonic shell, the outer prismatic layer is tile-shaped from the initiation of the accretionary growth. However, this structure has never been observed during post-embryonic stages of any Lower Jurassic species so far studied, although many (if not all?) postTriassic species display the same reticulate pattern of embryonic shell sculpture (see Chirat 2001). This suggests that rise of tile-shaped lamellae of the outer prismatic layer during post-embryonic stages similar to that of Cymatoceras and Cymatonautilus , could be not explained by terminal addition to the ancestral ontogeny, but rather, by heterochronic shifting of the embryonic growth pattern trough post-embryonic de- 64 R. Chirat & H. Bucher velopment (paedomorphosis). This hypothesis could provide a different perspective on the developmental processes involved in the evolution of Nautilida than do observations of the terminal addition alone. However, as a result of phyletic alteration in developmental timing of ancestral features, heterochrony alone would not explain a more profound issue, that is, how this structure appeared in the first place. Acknowledgements.
/ This work was partly supported by the Swiss SNF (project 2100
/068061, H. Bucher). N. Podevigne (Université Claude Bernard Lyon 1) is thanked for the photographic work. References Arnold, J. M. & Carlson, B. A. 1986: Living Nautilus embryos: preliminary observations. Science 232 , 73
/76. Arnold, J. M. & Landman, N. H. 1992: Embryology of Nautilus : evidence for two modes of shell ontogeny. Journal of Cephalopod Biology 2 , Abstract, 1. Abouheif, E., Akam, M., Dickinson, W. J., Holland, P. W. H., Meyer, A., Patel, N. H., Raff, R. A., Roth, V. L. & Wray, G. A. 1997: Homology and developmental genes. Trends in Genetics 13 , 432
/433 Bardhan, S. & Halder, K. 2000: Sudden origin of ribbing in Jurassic Paracenoceras (Nautiloidea) and its bearing on the evolution of ribbing in post-Triassic nautiloids. Historical BioIogy 14 , 153
/168. Binckhorst, J. T. 1861: Monographie des gastropodes et des céphalopodes de la craie supérieur du Limbourg. Classe des céphalopodes. 83 pp/44 pp. Brussels & Maastricht. Bucher, H., Chirat, R., & Guex, J. 2003: Morphogenetic origin of radial lirae and mode of shell growth in Calliphylloceras (Jurassic Ammonoidea). Eclogae Geologicae Helvetiae 96 , 495
/502. Bucher, H., Landman, N. H., Guex, J. & Klofak, S. M. 1996: Mode and rate of growth in ammonoids. In Landman, N. H., Tanabe, K. & Davis, R. A. (eds): Ammonoid Paleobiology, 407
/461. Topics in Geobiology 13 . Chirat, R. 2001: Anomalies of embryonic shell growth in post-Triassic Nautilida. Paleobiology 27 , 485
/499. Chirat, R. & Boletzky, S. von 2003: Morphogenetic significance of the conchal furrow in nautiloids: evidence from early embryonic shell development of Jurassic Nautilida. Lethaia 36 , 161
/170. Collignon, M. 1949: Recherches sur les faunes albiennes de Madagascar. I: L’Albien d’Ambarimaninga. Annales du Service Géologique et Minier de Madagascar 16 , 1
/128. d’Orbigny, A. 1840: Paléontologie française. Terrains crétacés. I. Céphalopodes . 662 pp. Masson, Paris. d’Orbigny, A. 1850: Prodromme de paléontologie stratigraphique universelle des animaux mollusques et rayonnés . 394 pp. Masson, Paris. Douvillé, H. 1916: Les terrains secondaires dans le Massif du Moghara à l’est de l’Isthme de Suez, d’après les explorations de M. CouyatBarthoux. Mémoire de l’Académie des Sciences 54 , 1
/184. Dzik, J. 1984: Phylogeny of the Nautiloidea. Palaeontologia Polonica 45 , 1
/255. Hyatt, A. 1883: Genera of fossil cephalopods. Proceedings of the Boston Society of Natural History 22 , 253
/338. Hammer, Ø. & Bucher, H. 1999: Reaction-diffusion processes: application to the morphogenesis of ammonoid ornamentation. Geobios 32 , 841
/852. LETHAIA 39 (2006) Kobayashi, T. 1954: A new cymatoceratid from the Palaeogene of northern Kyushu in Japan. Japanese Journal of Geology and Geography 24 , 15
/21. Kummel, B. 1956: Post-Triassic nautiloid genera. Bulletin of the Museum of Comparative Zoology 114 , 324
/494. Kummel, B. 1964: Nautiloidea-Nautilida. In Moore, R. C. (ed.): Treatise on Invertebrate Paleontology. Part K. Mollusca 3 , K383
/ K466. Geological Society of America and University of Kansas Press, Boulder, Colorado and Lawrence, Kansas. Matsumoto, T. & Muramoto, K. 1983: Cretaceous nautiloids from Hokkaido
/II. Part. 2. Three nautiloid species from the Santonian and Campanian of Hokkaido. Transactions and Proceedings of the Palaeontological Society of Japan 130 , 85
/95. Matsumoto, T., Takahashi, T., Obata, I. & Futakami, M. 1984: Cretaceous nautiloids from Hokkaido
/IV. Part. 4. An interesting nautiloid species from the Cenomanian of Hokkaido. Transactions and Proceedings of the Palaeontological Society of Japan 133 , 288
/ 299. Schlüter, C. 1876: Cephalopoden der oberen deutschen Kreide II. Palaeontographica 24 , 123
/264. Seilacher, A. 1984: Constructional morphology of bivalves: evolutionary pathways in primary versus secondary soft-bottom dwellers. Palaeontology 27 , 207
/237. Spath, L. F. 1927: Revision of the Jurassic cephalopods fauna of the Kachh (Cutch). Memoirs of the Geological Survey of India, Palaeontologia Indica 9 , 1
/84. Tanabe, K. & Uchiyama, K. 1997: Development of the embryonic shell structure in Nautilus . The Veliger 40 , 203
/215. Tintant, H. 1969: Les ‘nautiles à côtes’ du Jurassique. Annales de Paléontologie (Invertébrés) 55 , 53
/96. Tintant, H. 1984: L’évolution du concept de genre: de la similitude à la parenté. Bulletin de la Socie´té géologique de France 7 , 573
/582. Tintant, H. 1987: Les Nautiles jurassiques d’Arabie saoudite. In Enay, R. (ed.): Le Jurassique d’Arabie Saoudite centrale , 67
/159. Géobios , Mémoire Spécial 9 . Tintant, H. 1989: Les avatars ontogénétiques et phylogénétiques de l’ornementation chez les nautiles post-triasiques. In David, B., Dommergues, J. -L., Chaline, J. & Laurin, B. (eds): Colloque international CNRS: Ontogenèse et Evolution, Dijon 1986. Mémoire Spécial Géobios 12 , 357
/367. Tintant, H. 1993: L’évolution itérative des nautiles post-triasiques. In Elmi, S., Mangold, C. & Almeras, Y. (eds): Troisième symposium international sur les céphalopodes actuels et fossiles, 359
/372. Mémoire Spécial Géobios 15 . Tintant, H. & Kabamba, M. 1983: Le nautile, fossile vivant ou forme cryptogène? Essai sur l’évolution et la classification des nautilacés. Bulletin de la Société zoologique de France 108 , 569
/579. Wagner, G. P. (ed.) 2001: The Character Concept in Evolutionary Biology. 622 pp. Academic Press, San Diego. Ward, P. D. 1987: The Natural History of Nautilus . 267pp. Allen and Unwin, Boston. Ward, P. D. 1988: Form and function of the Nautilus shell: some new perspectives. In Biggelar, P. (ed.): The Mollusca 11, 143
/165. Academic Press, New York. Ward, P. D. & Saunders, W. B. 1997: Allonautilus : a new genus of living nautiloid cephalopod and its bearing on phylogeny of the Nautilida. Journal of Paleontology 71 , 1054
/1064. Wray, G. A. & Abouheif, E. 1998: When is homology not homology? Current Opinion in Genetics & Development 8 , 675
/680. Wiedmann, J. 1960: Zur Systematik jungmesozoischer Nautiliden unter besonderer Berücksichtigung der ibersischen Nautilinae d’Orb. Palaeontographica Abt . A 115 , 144
/206.