Introduction

Phricodoceras is a homogeneous and unambiguously de fined group among the late Sinemurian and Pliensbachian ammonites. Although generally scarce, Phricodoceras has been actively collected and studied since the early nine teenth century because of its attractive and unusual tubercu late ornamental pattern. As a result, despite its rarity, it is discussed in a hundred or so publications. It is surprising therefore that its relationships and consequently its taxo nomic attribution should recently have been seriously questioned (Dommergues 1993, 2003; Dommergues and Meister 1999; Meister 2007; Dommergues et al. 2008) and finally reconsidered at the superfamily level (Edmunds et al. 2003; Meister et al. 2010, 2011; Blau and Meister 2011). This edifying late taxonomic revision illustrates the surpris ing immovability of questionable practices in ammonite taxonomy. The aim of this work is, first, to summarize the latest discoveries and their taxonomic implications; second, to recapitulate the main steps of the taxonomic practices in volving Phricodoceras since the early nineteenth century; and third, to examine the grounds for the major changes in the interpretation of the relationships of Phricodoceras among the Sinemurian and Pliensbachian ammonites. Spe cial attention is also paid to why the misleading phyletic hy pothesis by which the genus Phricodoceras was ascribed to the Eoderoceratoidea should have proved so resilient in the literature. The case of Phricodoceras is discussed here to exemplify what is a common bias in ammonite taxonomic practices. Taxonomic groupings grounded on some “over all resemblance” combined with stratigraphic control are usually evidence. Unfortunately, later this may become “coarse” evidence and/or may be found to be homeo morphy, as is shown here for Phricodoceras. The impor tance of transitional forms in convincingly defining the pri mary homologies is also clearly illustrated in the example studied. Thus, beyond Phricodoceras, the present work can be viewed as a “case study” and a possible source of ideas for ammonite taxonomy.

Institutional abbreviations.-UBGD, University of Burgundy.

Other abbreviations.-M, macroconch; m, microconch; t1, latero-umbilical position; t2, latero-ventral position; t3, peri siponal position; us, umbilical seam; vb, ventral band (see also Fig. 4).

Stratigraphic and geographic Settings

The stratigraphic and paleobiogeographic frameworks of the genus Phricodoceras have been extensively and accurately described by Meister (2007: figs. 12, 14, 16, 17). The results of that key work are summarized here and supplemented schematically by more recently published data (Figs. 1, 2). The stratigraphic range of Phricodoceras is objectively docu mented from the base of the Echioceras raricostatum Chrono zone (Crucilobiceras densinodulum Subchronozone) to the top of the Pleuroceras spinatum Chronozone (Pleuroceras hawskerense Subchronozone). In the Mediterranean Tethys the last Phricodoceras (Phricodoceras aff. cantaluppii Fantini Sestini, 1978) are associated with Emaciaticeras (Meister et al. 2010). The earliest representatives of the genus Phricodo ceras (Echioceras raricostatum Chronozone) belong to the group of Phricodoceras gr. taylori (Sowerby, 1826)—P. lamel losum (Orbigny, 1844). They exhibit from the outset all of the impressive diagnostic features of the genus. Convincingly, Phricodoceras roots among the genus Angulaticeras and A. (Angulaticeras) spinosus Meister, Schlögl, and Rakús, 2010, a recently discovered species with a Phricodoceras-like juve nile stage, comes from a condensed Carpathian fauna suggest ing a period from the Arietites bucklandi to the Caenisites turneri chronozones (Meister et al. 2010). The condensed con text of this unusual fossiliferous locality must be underlined because the sedimentary processes often associated with con densation can explain the presence of an episode that is usu ally missing at the regional level (e.g., long lasting submarine exposure and/or erosion) (Olóriz 2000; Cecca 2002; Olóriz and Villaseñor 2010). Even if an age somewhere in the Caeni sites turneri Chronozone is plausible for A. (A.) spinosus (Fig. 1), there remains an undocumented stratigraphic gap in Phricodoceras history corresponding approximately to the duration of the Asteroceras obtusum—Oxynoticeras oxynotum Chronozones. Fig. 1 shows that Phricodoceras is clearly the longest-surviving genus of the Family Schlotheimiidae. In point of fact, the genus durations have a propensity to increase throughout the history of the family, and this tendency appar ently peaks with Phricodoceras. Obviously, like all taxo nomic groupings, genera are partly subjective and their dura tion may be influenced by taxonomic practice, which widely depends on human perception. Thus the long duration of Phricodoceras obviously reflects the persistence of only a few but striking ornamental diagnostic traits (autapomorphies). On the contrary, such flagrant features are missing among the earliest representatives of the family and the genus diagnoses are clearly less constrained as a result. So comparisons of ge nus durations are perhaps weakly significant in evolutionary words.

Fig. 1.

Bio-chronostratigraphic framework of the six genera belonging to the Schlotheimiidae as this family is understood in the present paper. The ranges are referred to the standard chronostratigraphic scale (stages and chronozones) so that relevant global comparisons can be made. The proba ble age of Angulaticeras spinosus is starred. Radiochronologic ages of the stage boundaries from Ogg et al. (2008). The height of the chronozone boxes varies with the stage duration in Myr.

In paleobiogeographic terms Phricodoceras is a taxon chiefly known in the Mediterranean and NW European con fines of the Western Tethys (Fig. 2). Thus, of the just over one hundred publications featuring, to some extent, the genus Phricodoceras, 37 concern the Mediterranean faunas (includ ing the Pontides, Northern Turkey), 41 discuss the NWEuro pean faunas, and only 7 refer to other parts of the world. In fact, very few specimens are cited outside the Mediterranean Tethys, NW Europe and the Pontides (Fig. 2). Moreover, the specimens from Western North America, Northern Chile and the Eastern Himalayas are unconvincing or questionable. The only reliable representative of the genus Phricodoceras from outside the Mediterranean and NW European confines of the Western Tethys is a finely preserved specimen from the Timor area close to the Australian Tethyan margin (Krumbeck 1922). Ideally, it would be best to consider the stratigraphic sensu lato and sedimentological frameworks so as to counter balance this crude palaeobiogeographical data, which can yield a partly biased picture of reality. Unfortunately, though, the present synthesis is grounded on such heterogeneous liter ature that the consideration of stratigraphic and sedimento logical data is no more than an ideal. Nevertheless-as previ ously demonstrated for the Early Pliensbachian by Dom mergues et al. (2009: fig. 6)-despite a similar study effort (at least in terms of number of publications), the Mediterranean Tethys palaeobiodiversity is clearly richer than that of NW Europe, although it is still undersampled in comparision.

Fig. 2.

Schematic distribution of Angulaticeras spinosus Meister, Schlögl, and Rakús, 2010 and Phricodoceras at the global scale. The approximate bound ary between the NW European and Tethyan (Mediterranean) faunas is suggested by a dotted line. Paleogeographical reconstruction from Vrielynck and Bouysse (2001), modified.

More generally, the Mediterranean Tethys seems to be the only known sustained “hot spot” of Phricodoceras di versity. By contrast, only a few species related to the group of P. taylori sensu lato are known inNWEurope and almost all of the many specimens known in this area are associated with a brief dramatic acme in the lower part of the Uptonia jamesoni Chronozone. Paradoxically, the Phricodoceras are never common in the Mediterranean Tethys but both their taxonomic diversity and their morphological disparity remain persistently high in this area where the genus is recurrently observed from the Echioceras raricostatum Chronozone to the base of the Pleuroceras spinatum Chronozone (Figs. 1, 2). We must also emphasize that Angulaticeras (Angulaticeras) spinosus, a possible ances tor of Phricodoceras, is to date only known in the Mediter ranean Tethys (i.e., Austroalpine). In terms of diversity (i.e., comparison of the number of species during the Echio ceras raricostatum and Uptonia jamesoni chronozones) the Pontides area occupies an intermediate position between the Mediterranean Tethys and NW Europe.

In this paper, the binominal italicized names of chrono zones result from the policy of the journal that any names derivative of biological species should be written in this way.

Morphology, dimorphism, ontogeny, and adaptation

The diagnostic features of Phricodoceras and especially the “juvenile” ornamental features are very unusual for Early Ju rassic ammonites and the genus has always been regarded as forming both a highly distinctive and a homogeneous taxon. Even the most morphologically derived forms (e.g., tiny Late Pliensbachian microconchs or large Early Pliensbachian macroconchs) can be fairly easily attributed to the genus. As a result, the synonymy of the genus is limited to a single taxon (i.e., Hemiparinodiceras Géczy, 1959) and there is no subgenus to suggest possible groupings within the twenty or so nominal species. Despite its apparent homogeneity, the genus Phricodoceras is not a simple lineage but, as evi denced by Meister (2007: fig. 15), a clade with a rather com plex internal structure. The concept of “species complex” might be helpful in putting the clade topology into words. Even if the phenomenon tends to decrease with time, a usu ally obvious microconch (m)/macroconch (M) dimorphism characterizes the Phricodoceras as exemplified by the pair of nominal species P. taylori (m)—P. lamellosum (M) in Fig. 3. Dimorphism seems to have peaked in this group close to the base of the Early Pliensbachian in NW Europe and therefore in a palaeobiogeographical context suggesting a briefly suc cessful northward faunal ingression. The extent of this strik ing dimorphism is difficult to quantify because the largest known P. lamellosum (M) are all incomplete phragmocones (e.g., Fig. 3A), and their adult body chambers are unknown. However, a ratio of about one to ten in diameter can be rea sonably suspected. The intermediate and outer whorls of the large macroconch forms have rather involute and compres sed shells with slightly curved flanks and a rounded ventral area. The transition between the umbilical area and the base of the flanks is rounded without shoulders, although faint peri-siphonal shoulders (s3), inherited from juvenile peri siphonal tubercles (t3), may persist at relatively large dia meters (e.g., Fig. 3A, B). The ornamentation of crowded, fine, subdivided and slightly flexuous ribs is rather discreet and often somewhat irregular (e.g., Fig. 3A). At large diame ters the ribs may cross the ventral area. Thus, the pre-adult and probably also the adult (body chamber) habitus of the macroconch is coarsely comparable, at the same diameter, to that of Angulaticeras. Actually, at large diameters Phricodo ceras lamellosum (M) looks similar to Angulaticeras al though with a less compressed shell and a wider and more rounded ventral area. In contradistinction, the traits of the inner whorls of the macroconch and microconch at all onto genetic stages (e.g., P. taylori) are quite distinctive and pre clude any confusion. At small diameters Phricodoceras may display one of the most impressively tuberculate ornamenta tions among the Early Jurassic ammonites, notably an excep tional peri-siphonal (t3) row of tubercles or spines (Figs. 4, 5). The inner mould of the phragmocones exhibits only the bases of the spines, which in this case look like truncated tu bercles or bullae (Fig. 4), but some well-preserved speci mens display prominent spines especially in peri-siphonal (t3) and latero-ventral (t2) positions (e.g., Buckman 1911: pl. 33; Hoffmann 1982: pl. 14: 3; Edmunds et al. 2003: fig. 20.5) (Fig. 5E). Within the groups of P. taylori (m)—P. lamellosum (M) and of P. bettoni (m) Géczy, 1976—P. urkuticum (M) (Géczy, 1959) at least, up to three rows of tubercles can be observed, though briefly, during the most strongly orna mented growth stage (Meister 2007: fig. 11). The positions of these three rows of tubercles are indicated in Fig. 4. Among the genus Phricodoceras the latero-umbilical (t1) row of tubercles is often missing and the latero-ventral (t2) row is sometimes absent, even in the group of P. taylori (m)—P. lamellosum (M) (e.g., Phricodoceras aff. cornutum [Simpson, 1843]) (Fig. 3D). Conversely the peri-siphonal (t3) row of tubercles remains visible, during a brief growth stage at least. The permanence of this trait is strong evidence that the peri-siphonal tubercles or shoulder (t3 or s3) of Phri codoceras are homologous with the sudden peri-siphonal in terruption of the ribs or shoulders (s3) of Angulaticeras, which is also a very permanent juvenile trait (Figs. 4–6). Although less distinctive, the suture lines of Phricodoceras also have informative features which can be contrasted with Angulaticeras on the basis of a comparative study of septal suture ontogenies. The pointed, often slender and trifid (sometimes sub-triangular) lateral lobe of Phricodoceras is the most obvious similarity (Fig. 7), and despite many appar ent differences, the suture line of Phricodoceras can be un derstood as a simplified version of that observable in Angu laticeras with wider saddles and chiefly without any clear re tracted suspensive lobe, as is usual in Angulaticeras. Many of these differences and especially the lack of an obvious sus pensive lobe are probably partially correlated with different shell morphologies. At the same diameters, shells are clearly more involute and compressed in Angulaticeras than in Phri codoceras whose inner whorls, at least, often have sub circular sections and barely overlap the successive whorls, thus providing less space for the retraction of the umbilical lobes. Conversely, the suture lines of Phricodoceras are very different from those of both the Lytoceratoidea and Eodero ceratoidea whose bifid or trifid lateral lobes are invariably adapically broad but abapically often narrow (Fig. 8).

The evolution of Phricodoceras is, as demonstrated by Meister (2007: fig. 11), basically controlled by ontogenetic heterochronies in the “size-based” or “allometric” and not “age-based” sense of the term. Fig. 9 summarizes and sim plifies the model proposed byMeister (2007) for Phricodo ceras and extends it to a broader taxonomic framework including Angulaticeras, with A. boucaultianum (Orbigny, 1844) (Early Sinemurian) for comparison and A. spinosus (Late Sinemurian) as a possible ancestor or at least the sister group of Phricodoceras (Late Sinemurian to Late Pliens bachian). The first step (A. boucaultianum to A. spinosus) involves a “juvenile innovation” sensu Dommergues et al. (1986) and Dommergues (1987), a phenomenon that is not a heterochony sensu stricto but which immediately precedes evolutionary phenomena chiefly controlled by heterochro nies. In the case of A. spinosus, the innovation is the possi bly rapid emergence of an obviously tuberculated ornamen tation in the innermost whorls only. Conversely, the subse quent and merely ribbed growth stages of this species are usual for Angulaticeras. Truncated tubercles in (t2) posi tion are clearly visible up to an umbilical diameter of 11mm (Fig. 5A₁, A₂). They are similar to the tubercles in the same position and at the same diameter in Phricodoceras (Fig. 5B₁) so, and although the ventral area is concealed by whorl overlap, it is plausible that tubercles also exist in peri siphonal position in the inner whorls of A. spinosum. The second step (A. spinosus to P. lamellosum) is chiefly a paedomorpic pattern of heterochony with an obvious decel eration of growth sensu Reilley et al. (1997). As is often the case, retardation is accompanied by a dramatic enhance ment of the juvenile features and the tuberculated ornamen tation reaches a maximum in the group of P. taylori (m)—P. lamellosum (M). The spines reach outstanding proportions and three rows of tubercles are usual. The third (P. lamel losum to P. urkuticum) and fourth (P. urkuticum to P. paronai [Bettoni, 1900]) steps follow a reversal and an in crease in complexity of the heterochronic pattern. These last two steps in Phricodoceras history witness a sustained contraction and weakening of the juvenile tuberculate stage and a correlative progressive decline in adult size. This complex pattern suggests the combination of two distinct polarities, one peramorphic (by acceleration of growth) and the other paedomorphic (by hypomorphosis), although “phyletic dwarfism” is another possibility because size is not necessarily a proxy of age. In palaeobiogeographical terms the late tiny or at least smallish (possibly dwarf ?) Phricodoceras are rare, or even very rare, strictly Tethyan species; however, relations with the palaeoenvironmental conditions remains obscure.

Fig. 3.

Microconch (m) / macroconch (M) dimorphism expressed by scholtheimiid ammonoid Phricodoceras exemplified by the NW Europe forms in the Uptonia jamesoni to Tragophylloceras ibex chronozones. A. Phricodoceras lamellosum (Orbigny, 1844) (M), UBGD 277451, Mazenay, Saône et Loire, France, probably early Uptonia jamesoni Chronozone, in apertural (A1), lateral (A2), and ventral (A3) views. B. Phricodoceras lamellosum (M), Kircheim unter Teck, Baden-Würtemberg, Germany, Early Pliensbachian (from Schlegelmilch 1976: pl. 27: 4, modified; original from Quenstedt 1884: pl. 28: 24), in apertural (B₁), lateral (B₂), and ventral (B₃) views. C. Phricodoceras taylori (Sowerby, 1826) (m), Corbigny, Nièvre, France, Uptonia jamesoni Chronozone, Phricodoceras taylori Subchronozone (from Dommergues 2003: pl. 1: 4), in lateral view. D. Phricodoceras aff. cornutum (Simpson, 1843) (m), Fresnay le-Puceux, Calvados, France, Early Pliensbachian (from Dommergues et al. 2008: pl. 3: 6, modified), in ventral (D₁) and lateral (D₂) views. E. Phricodoceras taylori (m), Fresnay-le-Puceux, Calvados, France, Early Pliensbachian (from Dommergues et al. 2008: pl. 3: 5, modified), in ventral (E₁) and lateral (E₂) views. The two specimens corresponding to A, B are incomplete phragmocones (juvenile or immature shells) but the three corresponding to C–E are adult microconchs with the major part of the body chamber. The end of the phragmocone is starred. Notice the progressive ontogenetic transformation from tubercle (t3) to faint shoulder (s3) in specimen B. Abbreviations: t2, tubercle in latero-ventral position; t3, tubercle in peri-siphonal position; s3, shoulder peri-siphonal position.

Fig. 4.

Position and terminology of the tubercles, spines and/or bullae on Phricodoceras shells (juvenile and/or microconch. A, B. Normal view. C, D. Shaded view with indication of the main ornamental structure outlines (white lines). Abbreviations: t1, latero-umbilical position; t2, latero-ventral posi tion; t3, peri-siphonal position; us, umbilical seam; vb, ventral band.

In terms of adaptation and traits of life history only as sumptions are possible. Nevertheless, the importance of pat terns chiefly related with juvenile stages (i.e., juvenile inno vation and paedomorphosis by deceleration) suggests that the evolutionary history of Phricodoceras was a phenome non partly associated with changes in juvenile living condi tions (Fig. 10). It seems reasonable to assume that the spec tacular tuberculate ornamentation ensured an effective pas sive protection both for the juvenile macroconchs and for the microconchs throughout their growth. In this sense, the emergence of a tuberculate growth stage in Phricodoceras, and therefore within the Schlotheiimidae, could be under stood as a convergence with the plentiful and diversified Late Sinemurian and Early Pliensbachian tuberculated Eodero ceratoidea (Fig. 11B, C). Conversely, it is possible that the living conditions of the post-juvenile macroconchs of Phri codoceras were little changed from those of Angulaticeras. Differences in lifestyle between juvenile macroconchs and microconchs (assumed to have been not very mobile but pas sively protected) and adult macroconchs (assumed to have had better hydrodynamic abilities and mobility, as suggested by the more compressed shell, with weaker and more flexu ous ornamentation) are therefore perhaps the key to the spe cific features of Phricodoceras. This hypothesis, summa rized in Fig. 10, is partly speculative, though, because eco ethological considerations derived from shell type and sculp ture with respect to “abilities” for swimming and/or maneu verability are interesting but unfortunately limited for all ectocochleate cephalopods (Westermann and Tsujita 1999).

Fig. 5.

Comparison of morphological and ornamental patterns of scholtheimiid ammonoid Angulaticeras spinosus Meister, Schlögl, and Rakús, 2010 and Phricodoceras gr. taylori (Sowerby, 1826) (m)—Phricodoceras lamellosum (Orbigny, 1844) (M). A. Angulaticeras (Angulaticeras) spinosus (M?), holotype, Chtelnica, Male KarpatyMts., Western Carpathians, Slovakia, Sinemurian condensed bed (fromMeister et al. 2010: fig. 34, a, b, modified), in lateral (A₁, A₂) and apertural (A₃) views. B. Phricodoceras taylori (m?), Corbigny, Nièvre, France, Uptonia jamesoni Chronozone, Phricodoceras taylori Subchronozone (from Dommergues 2003: pl. 1: 2, modified), in lateral (B₁, B₂) and ventral (B₃) views. C. Phricodoceras lamellosum (M), Hinterweiler, Baden-Würtemberg, Germany, Early Pliensbachian (from Schlatter 1980: pl. 6: 6,modified), incomplete phragmocone showing the transition between the juvenile tuberculate stage and the late merely ribbed stage, in lateral (C₁) and ventral (C₂) views.D. Phricodoceras taylori (m), Corbigny, Nièvre, France, Uptonia jamesoni Chronozone, Phricodoceras taylori Subchronozone (from Dommergues 2003: pl. 1: 4), in lateral view. E. Phricodoceras taylori (m), Fresnay-le-Puceux, Calvados, France, Early Pliensbachian (from Dommergues et al. 2008: pl. 3: 5, modified), in lateral (E₁) and ventral (E₂) views. To facilitate comparisons at small diameters, A₁ and B₁, respectively corresponding to A₂ and B₂, are twice magnified. The three specimens corresponding to A–C are incomplete phragmocones (juvenile or immature shells) but the two specimens corresponding to D, E are adult microconchs with the major part of the body chamber. The end of the phragmocone is indicated by a star. Some noticeable ornamental elements are indicated by arrows: smooth ventral band (vb), tubercle in latero-ventral position (t2), tubercle or shoulder in peri-siphonal position (t3 or s3).

Systematic palaeontology

The phylogenetic and taxonomic Quest

Phricodoceras in the literature.—Since 1826, a hundred or so publications have dealt, at least in part, with Phricodo ceras. Most of them contain illustrations (drawings or pho tographs). All these publications are considered in Fig. 12 with a view to summarizing the taxonomic opinions of their authors (Sowerby 1826; Zieten 1830; Orbigny 1844; Quen stedt 1846, 1849, 1883; Oppel 1853, 1856; Hauer 1861; Wright 1880; Fucini 1898, 1908; Bettoni 1900; Del Cam pana 1900; Hyatt 1900; Buckman 1911, 1921; Krumbeck 1922; Schröeder 1927; Höhne 1933; Gérard and Théry 1938; Roman 1938; Spath 1938; Otkun 1942; Venzo 1952; Fantini Sestini and Paganoni 1953; Donovan 1954; Arkell et al. 1957; Géczy 1959, 1979, 1998; Dean et al. 1961; Fantini Sestini 1962, 1978; Schindewolf 1962; Bremer 1965; Cantaluppi and Brambillia 1968; Frebold 1970; Wiedmann 1970; Tintant et al. 1975; Schlegelmilch 1976; Schlatter 1977, 1980, 1990, 1991; Dommergues 1978, 1993, 2003; Dubar and Mouterde 1978; Alkaya 1979; Linares et al. 1979; Wiedenmayer 1980; Donovan et al. 1981; Hoffmann 1982; Venturi 1982; Braga 1983; Mouterde et al. 1983; Büchner et al. 1986; Meister and Sciau 1988; Smith et al. 1988; Dommergues and Meister 1990, 1999; Dommergues et al. 1990, 2000, 2008; Cope 1991; Ferretti 1991; Sciau 1991; Tipper et al. 1991; Page 1993, 2008; Dommergues and Mouterde 1994; Mouterde and Dommergues 1994, Alkaya andMeister 1995; El Hariri et al. 1996; Faraoni et al. 1996; Smith and Tipper 1996; Géczy and Meister 1998, 2007; Rakús 1999; Macchioni 2001; Venturi and Ferri 2001; Howarth 2002; Rakús and Guex 2002; Donovan and Surlyk 2003; Edmunds et al. 2003; Meister et al. 2003, 2010, 2011; Hillebrandt 2006; Meister 2007; Yin et al. 2007; Venturi and Bilotta 2008; Venturi et al. 2010; Blau and Meister 2011).

In all, 162 specimens are figured in these publications, in cluding 78 for NW Europe and 84 for the Tethyan realm sensu lato. Compared with other taxa, such a large number of illustrations is not in proportion to the relative scarcity of Phricodoceras in the fossil record but partly reflects the spe cial interest shown by authors in this morphologically aston ishing and taxonomically challenging group. In fact, the il lustrated specimens correspond to a significant portion of the samples collected in the NW European faunas and encom pass almost all of the samples recovered in Tethyan sensu lato areas. In this context, the literature is probably very rep resentative of the material collected over some two centuries, and largely housed in museums.

Fig. 6.

Habitus of some specimens belonging to scholtheimiid ammonoid Agulaticeras, the genus which represents the root of Phricodoceras. A. Angulaticeras (Sulciferites) charmassei (Orbigny, 1844), Stuttgart-Vaihingen, Baden-Würtemberg, Germany, Arietites bucklandi Chronozone, Coroniceras rotiforme Subchronozone (from Bloos 1988: pl. 11, modified), in lateral (A₁) and apertural (A₂) views. B. Angulaticeras (Boucaulticeras) boucaultianum (Orbigny, 1844), Chtelnica, Male Karpaty Mts., Western Carpathians, Slovakia, Sinemurian condensed bed (from Meister et al. 2010: fig. 42f, g, modified), in lateral (B₁) and ventral (B₂) views. C. Angulaticeras (Boucaulticeras) gr. deletum (Canavari, 1882), Jbel Bou Hamid, Central Hight At las (Rich), Morocco, Late Sinemurian (from Guex et al. 2008: pl. 4: 6, modified), in apertural (C₁) and lateral (C₂) views. D. Angulaticeras (Boucaulticeras) gr. rumpens (Oppel, 1862), Chtelnica, Male Karpaty Mts., Western Carpathians, Slovakia, Sinemurian condensed bed (from Meister et al. 2010: fig. 40c, d, modified), in ventral (D₁) and lateral (D₂) views. E. Angulaticeras (Sulciferites) chtelnicaense Meister, Schlögl, and Rakus, 2010, holotype, Chtelnica, Male Karpaty Mts., Western Carpathians, Slovakia, Sinemurian condensed bed (from Meister et al. 2010: fig. 32d, e, modified), in ventral (E₁) and lateral (E₂) views. A, C (and possibly B) are incomplete phragmocones (juvenile or immature shells) but the two specimens corresponding to D, E have a signifi cant part of the body chamber intact. The age of D is doubtful but E is probably an adult. The end of the phragmocone is indicated by a star. The ornamenta tion of Angulaticeras is chiefly constituted by usually crowded, fairly flexuous and divided ribs which suddenly break up just before reaching the venter. At least at small diameters (juveniles, microconchs) the ventral area bears a narrow smooth and more or less depressed ventral band (vb). The abrupt endings of the ribs look like shoulders in peri-umbilical position (s3). Shoulders may vanish progressively with growth (B). Moreover, some rare species may exhibit unusual peri-umbilical projections from the ribs (ppr), which partially obstruct the umbilicus (E). Such projections are not true tubercles or spines.

Hypotheses, discussions, and facts.—From Sowerby (1826) to Hauer (1861), the early authors described and depicted some convincing specimens belonging to the group of Phri codoceras taylori under the generic name Ammonites without any indication of possible relationships within this huge genus (Fig. 12). Publications during the subsequent period from Wright (1880) to Del Campana (1900) still lack explicit infor mation about the possible relationships of the Phricodoceras at the family level. Nevertheless, the arrangement of the illus trated specimens on the plates (e.g., Quenstedt 1883–1885) and/or the use of genus names such as Aegoceras or Dero ceras (e.g., Wright 1880; Bettoni 1900) suggest that the au thors suspected possible relationships with certain taxa cur rently attributed to the Eoderoceratoidea (e.g., Liparocera tiadae). This pre-family position is clearly supported by the presence of tubercles and/or spines. At that same time, Hyatt (1900: 586–587) proposed the genus name Phricodoceras. Curiously this author included his new taxon in the “Cosmo ceratidae” family with some Middle Jurassic forms (i.e., Kos moceras and Sigaloceras) and surprisingly, at an informal higher taxonomic level, in the “Cosmoceratida” with some Cretaceous taxa (e.g., Douvillieiceras). The grouping at fam ily level proposed by Hyatt (1900) is based on obvious orna mental convergences and it is currently rejected as strongly polyphyletic. Only Gérard and Théry (1938) followed Hyatt's (1900) proposal. On the contrary, Buckman (1911, 1921) ex plicitly includes Phricodoceras within the Liparoceratidae thereby clarifying and formalizing the implicit hypothesis of many previous authors. From that time until fairly recently- even if Spath (1938) creates the subfamily Phricodoceratinae (within the Eoderoceratidae)-Phricodoceras was under stood, usually unreservedly, as belonging to the Eoderocera toidea. The single notable exception is Wiedmann (1970: 1002) who proposes that Phricodoceras is a possible relative of Adnethiceras within the Lytoceratoidea. In fact, at the superfamily level, the authors tend to conform to the position of Arkell et al. (1957) even if the family and subfamily levels are sometimes challenged. For example, the grouping of Phri codoceras and Epideroceras within the Phricodoceratinae proposed by Arkell et al. (1957) is abandoned by several au thors (e.g., Cope 1991; Schlatter 1991; Dommergues and Meister 1999). Nevertheless, it was not until 1991 that the in clusion of Phricodoceras in the Eoderoceratoidea was seri ously challenged by Kevin Page (personal communication to Dommergues 1993) and that convincing relationships with the Schlotheimiidae within the Psiloceratoidea were considered for the first time to be at least a plausible hypothesis. Despite this first serious challenge to the traditional taxonomic attribu tion, most authors until Yin et al. (2007) continued to consider Phricodoceras as member of Eoderoceratoidea with no fur ther discussion. In spite of this taxonomic inertia, several pub lications have understood Phricodoceras as an unresolved taxon and two to four credible but rival hypothesis have been suggested (Dommergues 1993, 2003; Dommergues and Meister 1999; Meister 2007; Dommergues et al. 2008). In all these papers, the possibility of the Schlotheimiidae and Phri codoceras being closely related is seriously considered but Edmunds et al. (2003) were clearly the first to propose this tax onomic option unreservedly albeit unfortunately without any compelling evidence. Later, Page (2008) took up this position but with some reservations. Such an alternative was discussed also by Venturi and Bilotta (2008) and Venturi et al. (2010), and their choice of a doubtful superfamily classification for the Phricodoceratidae was due to the lack of decisive data. The proof that Phricodoceras belongs to the Schlotheimiidae was ultimately provided by Meister et al. (2010), who described a new Angulaticeras (i.e., A. spinosus) whose inner whorls are virtually indistinguishable from those of Phricodoceras gr. taylori—P. lamellosum at the same diameter. Since this publi cation, all subsequent works have placed the Phricodoceras within the Psiloceratoidea and close to or within the Schlo theimiidae (Blau and Meister 2011; Meister et al. 2011).

Fig. 7.

Septal suture lines of several Schlotheimiidae belonging to the genera Phricodoceras (A–E) and Angulaticeras (F–J). A. Phricodoceras urkuticum (Géczy, 1959) (from Géczy 1976: fig. 49, modified). B. Phricodoceras taylori (Sowerby, 1826) (from Dommergues 2003: fig. 6A, modified). C. Phricodoceras taylori (from Dommergues 2003: fig. 6B, modified). D. Phricodoceras taylori (from Schlegelmilch 1976: 61, modified). E. Phricodo ceras gr. taylori (Sowerby, 1826) (from Schlatter 1990: fig. 3, modified). F. Angulaticeras martinischmidti (Lange, 1951) (from Schlegelmilch 1976: 38, modified). G. Angulaticeras charmassei (Orbigny, 1844) (from Schlegelmilch 1976: 38, modified). H. Angulaticeras densilobatum (Pompeckj, 1893) (from Schlegelmilch 1976: 39, modified). I. Angulaticeras lacunatum (J. Buckman, 1844) (from Schlegelmilch 1976, 38, modified). J. Angulaticeras rumpens (Oppel, 1862) (from Schlegelmilch 1976: 39, modified). For each suture line the whorl height (wh) is indicated, if known. The main elements of the suture line are indicated by following abbreviations: E, external lobe; L, lateral lobe; U₁, U₂, umbilical lobes; I, internal lobe.

Fig. 8.

Septal suture lines of several Lytoceratoidea (A–D) and Eoderoceratoidea (E–H) for comparisons with those of the scholtheimiid ammonoids Angulaticeras and Phricodoceras (Fig. 7). A. Zaghouanites arcanum (Wiedenmayer, 1977) (from Rakús and Guex 2002: fig. 54e, modified). B. Eolyto ceras tasekoi Frebold, 1967 (from Wiedmann 1970: text-fig. 9c, modified). C. Pleuroacanthites biformis (Sowerby in De La Beche, 1831) (from Canavari 1888: text—fig. 2.3, modified). D. Analytoceras gr. articulatum (Sowerby in De La Beche, 1831) (from Wiedmann 1970: text—fig. 8a, modified). E. Epi deroceras planarmatum (Quenstedt, 1856) (from Schlatter 1980: beil. 15a, modified). F. Xipheroceras rasinodum (Quenstedt, 1884) (from Schlegelmilch 1976: 57, modified). G. Xipheroceras ziphus (Zieten, 1830) (from Schlegelmilch 1976: 56, modified). H. Eoderoceras bisbinigerum (Buckman, 1918) (from Schlegelmilch 1992: 62, modified). For each suture line the whorl height (wh) is indicated, if known. The main elements of the suture line are indi cated by following abbreviations: E, external lobe; L, lateral lobe; U₁, U₂, umbilical lobes; I, internal lobe.

Fig. 9.

Some illustrative steps-in terms of morphological ontogeny-in the intricate evolutionary trend from the Sinemurian scholtheimiid genus Angulaticeras to the late Pliensbachian Phricodoceras (i.e., Phricodoceras paronai [Bettoni, 1900]). For simplicity, the complex and more or less gradual ontogenetic transformations are reduced to just four stages (see A–C for an illustration of the last three). The length and the place of a given stage in the ontogenetic cartouches depend on its duration and position during ontogeny. The overall length of the cartouche is proportional to adult size. Ontogenies of the macroconchs (M) alone are depicted in the cartouches and the adult sizes (complete shells) of the microconchs (m) are suggested by black triangles (grey if doubtful). A–C. Scholtheimiid ammonoid Phricododeras lamellosum (Orbigny, 1844), Rote Island, East Nusa Tenggara, Indonesia, probably Early Pliensbachian (from Krumbeck 1922: pl. 17: 5, modified), in ventral (A), lateral (B), and apertural (C) views.

Characters, assumed relationships, and taxonomic prac tice.-The history of taxonomic practice is rarely considered for itself, especially for ammonites (Donovan 1994). This is regrettable because such historical approaches may help to re fine taxonomic practices empirically by highlighting some misleading but consensual traditions. The case of Phricodo ceras is particularly instructive in this respect because a widely accepted hypothesis, herein rejected, has affected the taxonomic understanding of this remarkable group of am monites. This confusing but successful hypothesis is based on a dual argument grounded on both the concepts of “overall re semblance” and of “stratigraphic consistency”. Indeed, Phri codoceras and especially the emblematic P. taylori, which is locally not rare in the Uptonia jamesoni and Tragophylloceras ibex chronozones (Early Pliensbachian), can be roughly com pared with some Late Sinemurian and/or Early Pliensbachian Eoderoceratoidea (e.g., Eoderoceratidae, Polymorphitidae, Liparoceratidae). Some of these more of less markedly tuberculated forms have subplatycone, subplanorbicone or subsphaerocone shells with usually rounded and keelless ven tral areas. The habitus of such Early Pliensbachian Eodero ceratoidea (Fig. 11) are not very close to those of Phri codoceras (Figs. 3–5) (e.g., lack of peri-siphonal tubercles but usually presence of ventral secondary and intercalary ribs be tween the ventro-lateral rows of tubercles in Eoderoceratoidea but not in Phricodoceras), but all these forms are roughly co eval and the presence of tubercles and/or spines was long re garded as a diagnostic trait confined or pretty much so to the Eoderoceratoidea among the Pliensbachian ammonites. Con trariwise, Schlotheimiidae were understood until recently as forms unable to produce true tubercles and/or spines. Thus, in addition to the age (chiefly Early Pliensbachian), the presence of tubercles, the keelless smooth ventral area and the rather evolute juvenile coiling pattern were all used as arguments (taxonomic shoehorns) for placing Phricodoceras within the Eoderoceratoidea. This nearly universally or at least widely accepted argument is in fact circular. It was ultimately over turned by the recent discovery by Meister et al. (2010) of a clearly tuberculate juvenile growth stage in the innerwhorls of a typical Schlotheimiidae (i.e., Angulaticeras spinosus). From then on, it becomes easy to understand the genus Phricodo ceras as a close relative of Angulaticeras within the Schlo theimiidae and to fundamentally rethink the comparative anat omy of these forms. For example, it becomes possible to prove the peri-siphonal shoulders of the Schlotheimiidae are homol ogous with the peri-siphonal tubercles of Phricodoceras. In fact, the homologies (e.g., shell morphology, ornamentation, suture line if controlled by ontogenesis) with Angulaticeras are so numerous and obvious, throughout the growth stages, that it seems unnecessary to use a distinct subfamily or family level name to separate the two genera.

Conclusions

The history of taxonomic practice with respect to Phri codoceras is edifying because it clearly exemplifies the vul nerability of approaches based on “overall similarity” even if they are stratigraphicaly well constrained. In addition, it shows how much an allegedly consensus-based formaliza tion such as that proposed in the “Treatise of Invertebrate Pa leontology” (Arkell et al. 1957) may become sterilizing for taxonomic research. The present synthesis suggests that the understanding of relationships between ammonites, and par ticularly between clearly identified and distinct groups, de pends largely on the discovery of transitional forms and/or series in an acceptable stratigraphic context. If heterochroni cal processes, possibly associated with innovation, are in volved (as is the case for Phricodoceras), such transitional forms are often informative and easy to interpret in evolu tionary and phylogenetic terms. Unfortunately, intermediate forms between obviously distinct groups are usually very rare and localized. For example, Angulaticeras spinosus, the key species for the understanding of the relationship between Angulaticeras and Phricodoceras, is known by only three specimens (including the holotype) from a single but ex tremely rich fossiliferous locality in the western Carpathians, Slovakia (Meister et al. 2010). This locality has yielded sev eral thousand specimens including various Angulaticeras so Angulaticeras spinosus is obviously extremely rare. The sed imentary context is certainly important. For example, con densed deposits are probably particularly favorable for the search of transitional forms. Nevertheless, and despite the probable scarcity of many transitional forms, field studies still appear to be the most reliable way to resolve many enig matic taxonomic problems and to clarify our knowledge of palaeobiodiversity. In the absence of intermediate forms and/or series, cladistic analysis can be a useful approach in attempting to reconstruct phylogenies, but frequent homo plasies and the weakness of many primary homologies in the absence of transitional forms mean that this type of approach is often quite frustrating and nothing can replace the discov ery of a key intermediate form.

Fig. 10.

Schematic representation and comparison of the ontogenies of an Angulaticeras macroconch (A. boucaultianum) and of a Phricodoceras macroconch (P. lamellosum) in a simplified diagram taking into account the assumed mobility (x-axis) and the assumed passive shell protection (y-axis). These parameters cannot be fully expressed quantitatively. Mobility de pends mainly on hydrodynamic abilities, which are correlated with shell ge ometry but also with some aspects of ornamentation. Marked ornamental traits may play an important role. For example a keel or a ventral groove may increase the hydrodynamic stability of the shell and thereby facilitate mobil ity, but prominent tubercles and/or spines may significantly increase hydro dynamic drag thereby reducing mobility. Conversely the prominence of or namentation (chiefly of tubercles and/or spines) may be an effective passive protection against predators. Although highly schematic and hypothetical, such a diagram can be understood as an approximate representation of an “adaptative landscape” in which successive growth stages can be roughly situated. This “adaptative landscape” can be divided into four quadrants la beled A–D. The two studied species occupy only quadrants A (rather poor mobility but good passive shell protection) and C (good mobility but poor passive shell protection). In fact, only the juvenile growth stages of Phri codoceras lamellosum are situated in quadrant A but all the other growth stages, of both species, are in quadrant C. This pattern underlines the adaptative peculiarity of the juvenile growth stages of Phricodoceras.

Fig. 11.

Habitus of some nodded, spined and/or tuberculate Lytoceratoidea (A) and Eoderoderatoidea (B–D). A. Analytoceras hermanni (Gümbel, 1861), Kammerkaralpe, Waidring, Tyrol, Austria, probably Late Hettangian (from Wähner 1894: pl. 3: 3a, b, modified), in ventral (A₁) and lateral (A₂) views. B. Epideroceras lorioli (Hug, 1899), St Peter's Field, Radstock, Somerset, UK, Echioceras raricostatum Chronozone, Paltechioceras aplanatum Sub chronozone (from Edmunds et al. 2003: fig. 21. 4, modified), in lateral (B₁) and apertural (B₂) views. C. Tetraspidoceras repentinum Edmunds, 2009, St Pe ter's Field, Radstock, Somerset, UK, Uptonia jamesoni Chronozone, Phricodoceras taylori Subchronozone (from Edmunds 2009: pl. 32: 1, modified), in lateral (C₁) and ventral (C₂) views. D. Becheiceras bechei (Sowerby, 1821), Golden Cap, Seatown, Dorset, UK, Prodactylioceras davoei Chronozone, Oistoceras figulinum Subchronozone (from Edmunds 2009: pl. 38: 1, modified), in lateral (D₁) and apertural (D₂) views. Tubercles and/or spines in (t1) and/or (t2) positions of the Eoderoceratoidea (B–D) are not homologous with those of Phricodoceras, nevertheless this genus was long understood as a (borderline) member of this superfamily. In the case of Lytoceratoidea (A) the ornamental features in peri-siphonal position (pn3) are parabolic nodes which are morphologically clearly distinct from the tubercles or spines of both Eoderoceratoidea and Phricodoceras. The growth stage of the specimen is un known.

Fig. 12.

Historical synthesis of the taxonomic interpretation for the genus Phricodoceras from 1826 until today. Six options are considered: H?, no taxo nomic attribution or attribution deliberately left undetermined; Eo, explicit attribution to Eoderoceratoidea or implicit proximity with some ammonites cur rently attributed to the Eoderoceratidae; Ko, explicit attribution to the Kosmoceratidae; Ly, explicit attribution to the Lytoceratoidea (in the current sense); Ps, explicit attribution to the Psiloceratoidea and proximity with the Schlotheimiidae; La, enigmatic lazarus taxon. A cross indicates an absence of attribu tion to a taxon. A single black dot suggests an implicit or explicit but very reserved attribution. Two black dots suggest an explicit but debatable attribution. Three black dots suggest an unconditional explicit attribution. Four black dots suggest an explicit attribution based on ontogenetic evidence. For easy read ing, the two columns corresponding to the two most frequent taxonomic interpretations (i.e., Eo and Ps) are shaded.