On the Rarity and Peculiarity of the Early Toarcian (Lower Jurassic) Leukadiellinae Ammonites—Systematic Review and Insights on the Interplay of Environmental Stress, Evolution and Biodiversity

: The subfamily Leukadiellinae Macchioni and Venturi, 2000 includes the two rare genera Leukadiella Renz, 1913 and Renziceras Arkell, 1953. Genus Leukadiella is characterised by ornamental and structural features unusual to the family Hildoceratidae Hyatt, 1867; for this reason, it has been occasionally grouped with other “odd” representatives of this family, as, for instance, Frechiella Prinz, 1904 and Paroniceras Bonarelli, 1893. While Renziceras comprises only one species, the eleven species currently recognised within Leukadiella show a noticeably discontinuous variability ﬁeld; together with rarity of ﬁndings, the interpretation of this variability represents a renowned obstacle to their classiﬁcation and phylogenetic reconstruction. Following an analysis of the morphological characters commonly used for deﬁning Leukadiella species, a revised taxonomic scheme is herein proposed. Two clusters of characters are deﬁned, based on which two morphologically distinct groups of species are distinguished, referred to as Helenae Group and Ionica Group. The genus name Leukadiella is maintained for species of the Helenae Group, which are restricted via synonymy to Leukadiella helenae Renz, 1913 and Leukadiella jeanneti Renz, 1927. The new genus name Neoleukadiella gen. nov. is proposed for species of the Ionica Group, which, by analogous limitation, are Neoleukadiella ionica Renz and Renz, 1946 and Neoleukadiella gallitellii Pinna, 1965. The reduction in the number of species from eleven to four is consistent with their rarity and limited paleogeographic distribution. The closely related Renziceras is considered the direct progenitor of Leukadiella ; in turn, the Apennine genus Cingolites Sassaroli and Venturi, 2010 is proposed as transitional between Hildaites Buckman, 1921 and Renziceras . Conversely, the progenitor of Neoleukadiella remains uncertain, although a possible relationship between Leukadiella and Neoleukadiella species is tentatively traced. Finally, some challenging assumptions are made by emphasizing the role of environmental stress in controlling developmental dynamics that may drive striking phenotypic modiﬁcation, of the kind observed in the Leukadiellinae species.

These scenarios may be key for understanding the occurrence of rare and "strange" species in the fossil record, as in the case of Leukadiella and other early Toarcian groups. Because phenotypic novelties stemming during events of environmental stress may be significant [54], the post-crisis biodiversity is also affected, creating the premise for compensating (on the long term) the reduction in biodiversity caused by extinctions during the crisis acme. The late Pliensbachian-early Toarcian crisis is of interest for the subsequent renewal of ammonite species in greater part stemming from the Hildaites group [26,55], which coincides with the emergence of Leukadiella or of its closest precursor.

Historical Background
Despite being considered typical of Italy and Greece, Leukadiella species are rare in the Toarcian fauna of the Apennines, Southern Alps, and Greek area. The closest related genus Renziceras, with its only species R. nausikaae, is even less frequent than Leukadiella, with only 9 specimens reported in the literature: 7 specimens from the Apennines have been reported by Macchioni and Venturi [17]; one dubious fragment from Spain was described as Leukadiella sp. by Braga et al. [56], and later re-interpreted as R. nausikaae by Macchioni and Venturi [17]. Whereas Renziceras is unknown outside the Mediterranean region, other findings of Leukadiella are known only from Canada [25]; in addition, one ambiguous designation is also reported from South America [57]. The specimens from Canada differ significantly from the Mediterranean ones because of their larger growth size, with average diameter of 80 mm and a maximum diameter of 155 mm [25].
The Leukadiellinae subfamily was established by Macchioni and Venturi [17] to group Leukadiella and Renziceras while separating them from other peculiar hildoceratids. In this view, R. nausikaae was interpreted as the founder and most likely ancestor of Leukadiella. Renziceras is represented by only one species based on a specimen from the Toarcian of Greece (Epirus). This was described as Hildoceras nausikaae by Renz [15] (p. 607, pl. 14, figure 4, and text figures 25, 25a) (herein refigured in Figure 1a), later assigned to genus Bouleiceras [52] (p. 486), and finally designated as the new genus Renziceras by Arkell [16] (p. 36). Both Renz [15] and Arkell [16] were impressed by the "dimorphic" morphology of Renziceras, consisting of a markedly different ornamentation between the early and adult stage whorls. The inner whorls appear coronate, with strong spaced-out ribs terminated on the ventral edge by prominent tubercles, whereas ornamentation on the last whorl abruptly changes into less thick, gently sinuous ribs without tubercles (Figure 1a). The pre-adult stage morphology of Renziceras is overall similar to that displayed by L. helenae and other similar species (Figure 1b-h); on the other hand, it is significantly different from the basic morphological pattern of L. ionica and similar species with thinner ribs and both periumbilical and ventral nodes or tubercles (Figure 1i-l). This posed the problem of the meaning of such neat differences among Leukadiella species [17].
As in the case of Bouleiceras, the hypothesis of a direct derivation from the "tubercled" Hildaites is rooted in the supposed homology between the paired-sigmoidal ribs in Hildaites propeserpentinus [19] (Figure 3a) and in L. ionica (Figure 1i), thus conceived as the founder of the genus. Alternatively, authors assuming a coarse-clavate ribbed species as the founder of the genus [17] have proposed as indirect ancestors (via Renziceras) some coarse-ribbed Hildaites and Orthildaites ( Figure 3b); even earlier, also genera with a markedly sulcated-carenated venter-like Mercaticeras have been considered as possible ancestors [2,16].  [19]), considered the ancestor of L. ionica; (b) = Orthildaites douvillei Haug, 1884 [68] (holotype, from Gabilly [69]), considered a possible ancestor of L. helenae via R. nausikaae. Note that H. propeserpentinus is characterised by coupled ribs and umbilical nodes interpreted as precursors of the ornamental pattern in L. ionica ( Figure 1i); however, it lacks the ventral nodes. Also note that in O. douvillei the ontogenetic stages are reversed compared to R. nausikaae (Figure 1a), with the thicker ribbed stage in the outer whorls of Orthildaites and in the inner ones of Renziceras. Moreover, the thicker ribs of Orthildaites attenuate towards the venter instead of further thickening into a clavi, as in Renziceras and Leukadiella. According to Wendt [23], two branches stem from the founder species L. helenae ( Figure  4a): one leading to L. attenuata ( Figure 1g) and one along which species developed coupled ribs through "proterogenesis". This second branch is represented by the sequence: L. amuratica-L. ionica-L. sima (Figure 1f,i,l, respectively), the latter characterized by the simplification of ornamentation and the reoccurrence of single ribs only (Figure 1l). Macchioni and  [24]. (b) = Bouleiceras nitescens Thévenin, 1906 [18] (syntype, from Muséum National d'Histoire Naturelle, Paris, http://coldb.mnhn.fr/ catalognumber/mnhn/f/r01517). Note the double (umbilical and ventral) row of tubercles in the inner whorls of Bouleiceras. Some authors consider these as precursors of the umbilical and ventral nodes in L. ionica (see Figure 1i).
As in the case of Bouleiceras, the hypothesis of a direct derivation from the "tubercled" Hildaites is rooted in the supposed homology between the paired-sigmoidal ribs in Hildaites propeserpentinus [19] (Figure 3a) and in L. ionica (Figure 1i), thus conceived as the founder of the genus. Alternatively, authors assuming a coarse-clavate ribbed species as the founder of the genus [17] have proposed as indirect ancestors (via Renziceras) some coarse-ribbed Hildaites and Orthildaites ( Figure 3b); even earlier, also genera with a markedly sulcatedcarenated venter-like Mercaticeras have been considered as possible ancestors [2,16].  [24]. (b) = Bouleiceras nitescens Thévenin, 1906 [18] (syntype, from Muséum National d'Histoire Naturelle, Paris, http://coldb.mnhn.fr/catalognumber/mnhn/f/r01517). Note the double (umbilical and ventral) row of tubercles in the inner whorls of Bouleiceras. Some authors consider these as precursors of the umbilical and ventral nodes in L. ionica (see Figure 1i).
As in the case of Bouleiceras, the hypothesis of a direct derivation from the "tubercled" Hildaites is rooted in the supposed homology between the paired-sigmoidal ribs in Hildaites propeserpentinus [19] (Figure 3a) and in L. ionica (Figure 1i), thus conceived as the founder of the genus. Alternatively, authors assuming a coarse-clavate ribbed species as the founder of the genus [17] have proposed as indirect ancestors (via Renziceras) some coarse-ribbed Hildaites and Orthildaites ( Figure 3b); even earlier, also genera with a markedly sulcated-carenated venter-like Mercaticeras have been considered as possible ancestors [2,16].  [19]), considered the ancestor of L. ionica; (b) = Orthildaites douvillei Haug, 1884 [68] (holotype, from Gabilly [69]), considered a possible ancestor of L. helenae via R. nausikaae. Note that H. propeserpentinus is characterised by coupled ribs and umbilical nodes interpreted as precursors of the ornamental pattern in L. ionica ( Figure 1i); however, it lacks the ventral nodes. Also note that in O. douvillei the ontogenetic stages are reversed compared to R. nausikaae (Figure 1a), with the thicker ribbed stage in the outer whorls of Orthildaites and in the inner ones of Renziceras. Moreover, the thicker ribs of Orthildaites attenuate towards the venter instead of further thickening into a clavi, as in Renziceras and Leukadiella.
According to Wendt [23], two branches stem from the founder species L. helenae ( Figure  4a): one leading to L. attenuata ( Figure 1g) and one along which species developed coupled ribs through "proterogenesis". This second branch is represented by the sequence: L. amuratica-L. ionica-L. sima (Figure 1f,i,l, respectively), the latter characterized by the simplification of ornamentation and the reoccurrence of single ribs only (Figure 1l). Macchioni and  [19]), considered the ancestor of L. ionica; (b) = Orthildaites douvillei Haug, 1884 [68] (holotype, from Gabilly [69]), considered a possible ancestor of L. helenae via R. nausikaae. Note that H. propeserpentinus is characterised by coupled ribs and umbilical nodes interpreted as precursors of the ornamental pattern in L. ionica ( Figure 1i); however, it lacks the ventral nodes. Also note that in O. douvillei the ontogenetic stages are reversed compared to R. nausikaae (Figure 1a), with the thicker ribbed stage in the outer whorls of Orthildaites and in the inner ones of Renziceras. Moreover, the thicker ribs of Orthildaites attenuate towards the venter instead of further thickening into a clavi, as in Renziceras and Leukadiella.
It is worth noticing that the coarse-ribbed Hildaites stock is also ancestral to the Apennine genus Cingolites, which, according to Sassaroli and Venturi [26] (p. 114], likely stemmed from Hildaites crassus Guex, 1973 [70]. Cingolites differs from Hildaites because of the overall clavate ribs producing the broad coronate venter with wide sulci (distinctive also with respect to Orthildaites), and the trapezoidal-quadrate whorl section ( Figure 5). On this basis, Cingolites is herein proposed (in addition to Bouleiceras, Hildaites, Orthildaites or even Mercaticeras) as a credible progenitor bridging Hildaites with the Renziceras-L. helenae line. This interpretation further reduces the morphological gap between the Hildaites and the clavate-ribbed Leukadiella species. Instead, the origin of non-clavate species, as for instance L. ionica, remains more problematic, essentially because: a) the It is worth noticing that the coarse-ribbed Hildaites stock is also ancestral to the Apennine genus Cingolites, which, according to Sassaroli and Venturi [26] (p. 114), likely stemmed from Hildaites crassus Guex, 1973 [70]. Cingolites differs from Hildaites because of the overall clavate ribs producing the broad coronate venter with wide sulci (distinctive also with respect to Orthildaites), and the trapezoidal-quadrate whorl section ( Figure 5).  It is worth noticing that the coarse-ribbed Hildaites stock is also ancestral to the Apennine genus Cingolites, which, according to Sassaroli and Venturi [26] (p. 114], likely stemmed from Hildaites crassus Guex, 1973 [70]. Cingolites differs from Hildaites because of the overall clavate ribs producing the broad coronate venter with wide sulci (distinctive also with respect to Orthildaites), and the trapezoidal-quadrate whorl section ( Figure 5). On this basis, Cingolites is herein proposed (in addition to Bouleiceras, Hildaites, Orthildaites or even Mercaticeras) as a credible progenitor bridging Hildaites with the Renziceras-L. helenae line. This interpretation further reduces the morphological gap between the Hildaites and the clavate-ribbed Leukadiella species. Instead, the origin of non-clavate species, as for instance L. ionica, remains more problematic, essentially because: a) the Figure 5. Comparison of the ornamental pattern of C. clavatus ((a-c); (a,b) = holotype; from Sassaroli and Venturi [26]) and R. nausikaae ((d), from Macchioni and Venturi [17]).
On this basis, Cingolites is herein proposed (in addition to Bouleiceras, Hildaites, Orthildaites or even Mercaticeras) as a credible progenitor bridging Hildaites with the Renziceras-L. helenae line. This interpretation further reduces the morphological gap between the Hildaites and the clavate-ribbed Leukadiella species. Instead, the origin of non-clavate species, as for instance L. ionica, remains more problematic, essentially because: (a) the transition amuratica-ionica proposed by Wendt [23] is based on very few specimens, and (b) the differences between the ornamentation of L. ionica and that of any other lower Toarcian species, including H. propeserpentinus, are even more striking and discontinuous. Table 1. List of Leukadiella specimens and species figured or described in the literature. Only specimens figured by each author are included in the total count of known specimens; for instance, Wendt [22] includes three L. ionica in the list of the available material without figuring them. When a specimen is refigured by diverse authors it is counted only once. For instance, Wendt [23] reviews and refigures specimens already figured by Renz [15,49,51,52,71], Renz and Renz [50] and Kottek [22]; Pinna [53] establishes L. gallitellii on a specimen already figured by Gallitelli [72]; Gallitelli-Wendt [73] figures a specimen of L. ionica already figured by Wendt [23].

Author(s) Number of Specimens Number of Species Findings from Italian Localities
Renz [13] 1 In addition to literature data, only 2 specimens of L. helenae are the result of more than three decades of sampling of the Rosso Ammonitico of the Central Apennines conducted by the author and other research groups at Sapienza University. Both specimens are from the same locality, near the town of Polino ( Figure 6). In addition to literature data, only 2 specimens of L. helenae are the result of more than three decades of sampling of the Rosso Ammonitico of the Central Apennines conducted by the author and other research groups at Sapienza University. Both specimens are from the same locality, near the town of Polino ( Figure 6). (c) = lithological and biostratigraphic scheme. Below the Bifrons zone, a sterile interval ("unnamed zone") extends to the base of the section, encompassing the stratigraphic interval usually assigned to the Serpentinum and Tenuicostatum zones [43,81].
Specimen PO_LE10G2 is a broken phragmocone with part of the body chamber, ca. 26 mm in diameter and severely eroded on both sides; specimen PO_LEd1 is a well-preserved individual only 18 mm in diameter, including greater part of the body chamber ( Figure 7). Therefore, the main data on which this review is based are the original illustration of holotypes and diagnosis by the designating author, or even later revisions of the original diagnosis [17,23]. It is worth noting that 29 among the 70 specimens documented in the literature (41% of the total) are from Italian localities; of these, 17 specimens have been described by Macchioni and Venturi [17], all collected in the Central Apennines and grouped within 6 different species (Table 1). Of the other 10 specimens (representing 5 species) reported in the literature from Italian localities, 5 are from the Central Apennines, 2 from the Southern Alps, and 3 from Rocca Busambra, in Sicily ( Table 1)  Specimen PO_LE10G2 is a broken phragmocone with part of the body chamber, ca. 26 mm in diameter and severely eroded on both sides; specimen PO_LEd1 is a wellpreserved individual only 18 mm in diameter, including greater part of the body chamber ( Figure 7). Therefore, the main data on which this review is based are the original illustration of holotypes and diagnosis by the designating author, or even later revisions of the original diagnosis [17,23]. It is worth noting that 29 among the 70 specimens documented in the literature (41% of the total) are from Italian localities; of these, 17 specimens have been described by Macchioni and Venturi [17], all collected in the Central Apennines and grouped within 6 different species (Table 1).
Of the other 10 specimens (representing 5 species) reported in the literature from Italian localities, 5 are from the Central Apennines, 2 from the Southern Alps, and 3 from Rocca Busambra, in Sicily ( Table 1) [82]. The exception to a bio-stratigraphic distribution limited to the Bifrons zone is one specimen that Macchioni and Venturi [17] assign to the final part of the Serpentinum Zone, in the same interval where they record the occurrence of R. nausikaae.

Classification Methods
Taxonomic characters and criteria for the classification of Leukadiella species and the reconstruction of their phylogenetic relationships are revised essentially based on literature data. The list of taxonomic characters most frequently used in the literature for the definition of Leukadiella species and their variability can be summarized as follows: whorl embracement: from evolute to moderately involute; whorl section shape: sub-quadrate or sub-trapezoidal or sub-rectangular; shape of the flanks: more or less rounded; shape of the umbilical wall and ventri-lateral shoulder: rounded or sloping; ventral keel and sulci: variably pronounced and variably carved, respectively; type and shape of the ribs: either simple or paired; paired ribs may merge only at umbilical nodes or even at ventral nodes (forming fibulae), with couplets sometimes formed by a more raised primary rib and an attenuated secondary one. The shape of the ribs is either thick-clavate and overall straight or thin and sinuous, at times also concave backward; number of ribs: less than 10-12 when coarse-clavate, up to more than 20 when thinner and sinuous; umbilical and/or ventro-lateral tubercles and nodes: variably pronounced and overall elongated; suture line: variably simplified, particularly the lobes.
A comparison of the classification schemes proposed by diverse authors indicates that the ornamental pattern provides the most noticeable taxonomic characters for the distinction of species. Based on the variability of ornamental characters described above, two main ornamental patterns can be distinguished, consisting in clavate ribs variably thick and well spaced-out throughout the whorl; or non-clavate ribs, overall thin and closely spaced. Other characters (e.g., whorl section shape; whorl embracement), as well as other aspects of the ornamentation (e.g., paired ribs, presence of umbilical and/or ventral tubercles and nodes) are subdued to the principal clavate/non-clavate ornamental pattern ( Figure 8).
distinction of species. Based on the variability of ornamental characters described above, two main ornamental patterns can be distinguished, consisting in clavate ribs variably thick and well spaced-out throughout the whorl; or non-clavate ribs, overall thin and closely spaced. Other characters (e.g., whorl section shape; whorl embracement), as well as other aspects of the ornamentation (e.g., paired ribs, presence of umbilical and/or ventral tubercles and nodes) are subdued to the principal clavate/non-clavate ornamental pattern ( Figure 8).

Figure 8.
Ornamental patterns and variability of rib type and shape: the two basic patterns are the clavate and spaced-out ribs or the non-clavate and denser ribs. Clavate ribs are variably thick, simple and straight, and more or less rounded (a,b). The occasional occurrence of periumbilical reliefs and intercalated secondary ribs may simulate paring ribs (c,d). Non-clavate ribs are simple or truly paired, either sinuous (e) or more backward concave (f). Simple ribs show both periumbilical and ventral tubercles (g); paired ribs show periumbilical nodes only (h) or both periumbilical and ventral nodes, forming fibulae (i).
This variability and its relationship with the currently accepted species of Leukadiella is summarized in Figure 9, with reference to the holotypes of each species and variety proposed by diverse authors (compare with Table 1).

Figure 8.
Ornamental patterns and variability of rib type and shape: the two basic patterns are the clavate and spaced-out ribs or the non-clavate and denser ribs. Clavate ribs are variably thick, simple and straight, and more or less rounded (a,b). The occasional occurrence of periumbilical reliefs and intercalated secondary ribs may simulate paring ribs (c,d). Non-clavate ribs are simple or truly paired, either sinuous (e) or more backward concave (f). Simple ribs show both periumbilical and ventral tubercles (g); paired ribs show periumbilical nodes only (h) or both periumbilical and ventral nodes, forming fibulae (i).
This variability and its relationship with the currently accepted species of Leukadiella is summarized in Figure 9, with reference to the holotypes of each species and variety proposed by diverse authors (compare with Table 1).
as other aspects of the ornamentation (e.g., paired ribs, presence of umbilical and/or ventral tubercles and nodes) are subdued to the principal clavate/non-clavate ornamental pattern ( Figure 8). This variability and its relationship with the currently accepted species of Leukadiella is summarized in Figure 9, with reference to the holotypes of each species and variety proposed by diverse authors (compare with Table 1).  Table 1) of the main taxonomic characters (see Figure 8). Besides the ornamental features, also whorl section shape is represented, roughly based on the width/height ratio (w/h ≤ 1 = compressed; w/h ≥ 1 = depressed). Note that species can be separated into two main morpho-groups: (1) clavate/depressed and (2) non-clavate compressed. In addition, two sub-groups can be recognized within each main group, namely: with thick or thin clavate ribs (the latter with faint umbilical tubercles or nodes); and with simple or paired non-clavate ribs (the latter also with ventral nodes).

Morphological Variability of Leukadiella Species and Classification Problems
A comparison of the currently accepted species of Leukadiella indicates that the main morphological characters and their variants (particularly those concerning the shape of ribs and the occurrence of nodes and tubercles) combine in order that two morphological groups can be defined: Morphological Group 1 (MG1): structural features such as moderately involute whorls and quadrate-depressed section are coupled with single, coarse-clavate, straight and spacedout ribs (usually no more than 10-12 per whorl) ending with ventral clavi (Figure 8a-d); periumbilical nodes and paired ribs, or even neat periumbilical tubercles, are absent ( Figure 9). Overall, this set of characters is typical of the following species (see Figure 1): L. helenae, L. ticinensis, L. reisi, L. jeanneti, L. amuratica, L. attenuata and L. morettinii.
Morphological Group 2 (MG2): overall evolute and quadrate-compressed whorls are associated with thin and denser (up to >20 per whorl) sigmoidal or backward concave ribs, ended by periumbilical and ventri-lateral tubercles or even welded at both ends by nodes to form fibulae (Figure 8e-i; see also Figure 9); overall, rib shape is inconstant (diverse patterns may be present in the same specimen) and irregular (ribs of the same type may differ in thickness and overall shape). This morphological pattern is shared by the following species (see Figure 1): L. ionica, L. gallitellii, L. lombardica and L. sima; it is also typical of L. ionica var. paganiensis, which, according to Wendt [23], is a junior synonymous of L. ionica.
The two distinct morphological groups corresponding to MG1 and MG2 are respectively referred to as Helenae Group and Ionica Group, with reference to the most representative species in each group; if and how these two groups can be bridged remains an open question. It is worth noting that the ventral clavi in L. helenae (and likely in other species within MG1) are remnants of protruding spines that rarely preserve [17] (pl. 4 figures 1, 2); therefore, the shape and prominence of the ventral clavi (as well as the degree of rib projection) may reflect differential preservation of spine remnants rather than interspecific morphological variants (see Figure 7e).
Three species within the Helenae Group share a peculiarity by which they subtly differ from the basic pattern of L. helenae; these are: L. jeanneti, L. amuratica and L. attenuata (Figure 1e-g, respectively; see also Figure 9). These are peculiar because of the occasional occurrence of a periumbilical relief on some ribs or because ribs tend to thin on the middle flank, as they stem from a thicker umbilical segment, somewhat resembling a periumbilical tubercle (particularly in the holotypes of L. jeanneti and L. attenuata; Figure 1e,g). Overall, L. jeanneti, L. amuratica and L. attenuata pose the following problem: on one hand, they appear similar to a point that they may be variants of the same species (in which case L. amuratica and L. attenuata would be junior synonymous of L. jeanneti); on the other, their morphological peculiarities (tendency towards rib thinning and paring, or even development of a periumbilical relief) recall features more clearly expressed in species of MG2 (e.g., compare the ornamental pattern in Figure 8c with that in Figure 8h; and that in Figure 8d with that in Figure 8g).
This twofold problem can be solved by considering the jeanneti-amuratica-attenuata morphotypes as bridging the two groups, although morphologically closer to the MG1 pattern ( Figure 9). This interpretation is consistent with the phylogenetic trend proposed by Wendt [23], according to which a sequence can be traced from L. helenae to L. ionica via L. amuratica (Figure 4a), along with a heterochronic process leading to the development of coupled ribs in the final species. In this view, the transition proposed by Wendt [23] can be supported by also involving L. attenuata and L. jeanneti ( Figure 10; compare with Figure 4a), either as distinct species or (more likely) as variants of the same species, which by priority would be L. jeanneti (Figure 1e). by Wendt [23], according to which a sequence can be traced from L. helenae to L. ionica via L. amuratica (Figure 4a), along with a heterochronic process leading to the development of coupled ribs in the final species. In this view, the transition proposed by Wendt [23] can be supported by also involving L. attenuata and L. jeanneti ( Figure 10; compare with Figure  4a), either as distinct species or (more likely) as variants of the same species, which by priority would be L. jeanneti (Figure 1e). Notwithstanding whether morphological data support the occurrence of the heterochronic process invoked by Wendt [23], this transition is key for bridging MG1 and MG2; alternatively, MG1 and MG2 can be considered as two distinct groups stemming separately from two different (though related) ancestors. Both the MG1 and MG2 groups are present in the Apennines and Mediterranean region, whereas the specimens from North America described by Jakobs [25] can be referred to MG2 only. In fact, the specimens that Jakobs assigns to L. amuratica [25] (figures 6.11, 6.15, 6.16) and L. aff. helenae [25] (figures 6.13, 6.14) do show a coarser ribbing pattern; yet the shape of the ribs and their pairing in umbilical and ventral nodes is typical of the Ionica Group. These specimens thus evidence a coarsening of the ribs in species of the Ionica Group, rather than the presence of the "helenae" ornament type (MG1) in the Leukadiella from North America.
The coarseness of the ornamentation in specimens from North America may depend on their larger growth size, reaching values (average diameter of 80 mm and a maximum diameter of 155 mm; Jakobs [25]) unusual to Leukadiella from the Apennines. Therefore, species of the Helenae and Ionica groups are markedly different not only in their basic morphology, but also because of their geographical distribution, with the Helenae Group exclusive of the Mediterranean area and the Ionica Group present also in North America, though with "giant" morphotypes if compared with those from the Mediterranean.

Restriction of the Number of Accepted Species
The eleven species of Leukadiella currently accepted have been originally defined mainly based on differences in the ornamental features. However, the low number of findings for each species does not afford confident determination of the variability field of taxonomic characters; moreover, the coarseness of ribs and tubercles or clavi is significantly biased by preservation. This is particularly relevant for species of the Helenae Group, in which the shape and largeness of the ventral clavi reflect the degree of preservation of the root of the spines after they break off [17]. An implication of this reasoning Notwithstanding whether morphological data support the occurrence of the heterochronic process invoked by Wendt [23], this transition is key for bridging MG1 and MG2; alternatively, MG1 and MG2 can be considered as two distinct groups stemming separately from two different (though related) ancestors. Both the MG1 and MG2 groups are present in the Apennines and Mediterranean region, whereas the specimens from North America described by Jakobs [25] can be referred to MG2 only. In fact, the specimens that Jakobs assigns to L. amuratica [25] (figures 6.11, 6.15, 6.16) and L. aff. helenae [25] (figures 6.13, 6.14) do show a coarser ribbing pattern; yet the shape of the ribs and their pairing in umbilical and ventral nodes is typical of the Ionica Group. These specimens thus evidence a coarsening of the ribs in species of the Ionica Group, rather than the presence of the "helenae" ornament type (MG1) in the Leukadiella from North America.
The coarseness of the ornamentation in specimens from North America may depend on their larger growth size, reaching values (average diameter of 80 mm and a maximum diameter of 155 mm; Jakobs [25]) unusual to Leukadiella from the Apennines. Therefore, species of the Helenae and Ionica groups are markedly different not only in their basic morphology, but also because of their geographical distribution, with the Helenae Group exclusive of the Mediterranean area and the Ionica Group present also in North America, though with "giant" morphotypes if compared with those from the Mediterranean.

Restriction of the Number of Accepted Species
The eleven species of Leukadiella currently accepted have been originally defined mainly based on differences in the ornamental features. However, the low number of findings for each species does not afford confident determination of the variability field of taxonomic characters; moreover, the coarseness of ribs and tubercles or clavi is significantly biased by preservation. This is particularly relevant for species of the Helenae Group, in which the shape and largeness of the ventral clavi reflect the degree of preservation of the root of the spines after they break off [17]. An implication of this reasoning is that the difficulty of defining the variability field of taxonomic characters and the overlooking of preservation bias may have determined an overestimation of the number of species.
With these considerations in mind, the morphological pattern displayed by L. helenae, L. ticinensis, L. reisi, and L. morettinii (Figure 1b-d,h) appears similar, and the differences on which their distinction has been based become less remarkable (Figure 9). Only in the case of L. jeanneti, L. amuratica and L. attenuata (Figure 1e-g), the presence of subtle periumbilical bumps or relative thinning of the ribs on the middle flank may represent some distinguishing features ( Figure 9). Nevertheless, even in this case the lack of a proper number of specimens is an obstacle to the confident evaluation of the taxonomic relevance of these features (herein evaluated only basing on the figure of the holotypes). With all the above considered, species of the Helenae Group such as L. ticinensis, L. reisi and L. morettinii (Figure 1c,d,h) are interpreted as junior synonymous of the senior L. helenae (Figure 1b,b1); similarly, L. amuratica and L. attenuata (Figure 1f,g) may be interpreted as junior synonymous of L. jeanneti (Figure 1e). In this view, the amuratica-attenuata-jeanneti ornamental pattern (described above) allows tentatively to separate L. jeanneti (Figure 1e) from L. helenae (Figure 1b,b1). The number of species of the Helenae Group would thus be reduced via synonymy to only two species: L. helenae and L. jeanneti.
With reference to the Ionica Group, the number of species as well seems overwhelming compared to the variety of taxonomic characters and their understanding in terms of ontogenetic/intraspecific variability. L. ionica (Figure 1i) apparently differs from the other species of the group (Figure 1j-l) because of the occasional occurrence of ribs paired by periumbilical and ventral nodes, forming fibulae ( Figure 9); these alternate with ribs pairing only at the periumbilical node, or even with simple ribs bending backward from an elongated umbilical tubercle. In contrast, fibulae are absent in L. gallitellii, L. lombardica and L. sima (Figure 1j-l), which are all comparable in showing only simple ribs, though retaining umbilical and ventral tubercles ( Figure 9). For this reason, L. gallitellii, L. lombardica and L. sima are likely the same species, as also remarked by Macchioni and Venturi [17]; accordingly, and by priority, this species should be named L. gallitellii. Considering that, apart from the holotype of L. ionica var. paganiensis [50], only one rather ambiguous exemplar has been figured [22] (see Table 1), this variety is herein considered synonymous to L. ionica, in accordance with Wendt [23]; therefore, the Ionica Group as well reduces to two species: L. ionica and L. gallitellii.

Revised Classification Scheme
The above results indicate that the combination of ornamental and structural features by which the Helenae Group and Ionica Group have been separated define a taxonomic rank above the species level. Accordingly, the differences between species of these two groups are interpreted as reflecting two distinct genera. Consequently, the name Leukadiella is maintained exclusively for the senior species L. helenae and the closely related species L. jeanneti. The new genus name Neoleukadiella gen. nov. is proposed for the only two species herein considered plausible among those listed in the Ionica Group: N. ionica and N. gallitellii, which thus result as a combinatio nova. Based on this taxonomic approach, the following genera and species are herein recognised within the stock of former Leukadiella species: L. helenae, L. jeanneti, N. ionica comb. nov. and N. gallitellii comb. nov. (Figure 1b,b1,e,i,j, respectively).
Regarding the higher rank classification, the former genus Leukadiella has been frequently grouped at subfamily rank with Frechiella and Paroniceras, essentially due to sharing a neat departure from the basic morphological pattern of the Hildoceratinae, rather than because of sharing any significant common character. Indeed, there is no morphological character by which any species of Leukadiella and Neoleukadiella gen. nov. can be grouped with species of Frechiella and Paroniceras below the family rank. The Leukadiellinae subfamily proposed by Macchioni and Venturi [17] is thus the more suitable option in which to place Leukadiella and Neoleukadiella gen. nov., together with Renziceras. In this view, Cingolites, a possible direct ancestor of Renziceras characterised by clavate ribs and a coronate ventral area, could be also included in the Leukadiellinae Subfamily. All the above considered, a revised classification scheme is herein proposed. Subclass Ammonoidea Zittel, 1884 [83] Order Ammonitida Hyatt, 1889 [84] Suborder Ammonitina Hyatt, 1889 [84] Superfamily Hildoceratoidea Hyatt, 1867 [14] Family Hildoceratidae Hyatt, 1867 [14] Subfamily Leukadiellinae Macchioni and Venturi, 2000 [17] Genus Leukadiella Renz, 1913 [15] Diagnosis (emended). Genus Leukadiella, as herein redefined, is characterized by moderately embracing whorls with quadrate-depressed section and a deep umbilicus; ribs are simple, straight, coarse-clavate, well spaced-out (about 7-10 per whorl) and terminate with a variably pronounced ventral clavi (Figure 1b-h, Figures 7 and 8a-d). Neat periumbilical tubercles or nodes are lacking, although in some specimens there is subtle evidence of occasional thickening or converging of ribs near the umbilicus. The venter is variably wide, also depending on the shape of the whorl, which in turn varies with size. The height of the central and lateral keels is also variable, as well as the depth/width of the sulci; the lateral keels are affected by the different preservation of the ventral clavi along the ribs, which are the inception of large spines that rarely preserve. All known specimens are small, the two largest ones reported in the literature reaching the diameter of 30 [17] and 31.5 mm [23]. The suture line shows elements (number and shape of saddles and lobes) comparable with the general pattern of the Hildoceratitic line, though simplified (see Macchioni and Venturi [17] for a detailed description of the suture line based on a relatively large data set).
Occurrence. Lower Toarcian, Bifrons Zone. Description. Moderately involute whorls and quadrate-depressed section, which becomes trapezoidal in correspondence of the ventral coarsening of ribs. Ribs are few and well-spaced, usually up to 10 per whorl or less. They are straight with a rounded or sharp axial crest (e.g., Figure 8a,b), and thicken regularly and rapidly on the middle flank, before culminating in broad clavi on the ventral margin. The clavi are the inception of unpreserved large spines; the umbilical segment of the ribs is instead narrow and deprived of tubercles, although in some ribs it may be thicker than in others (Figure 8c,d). The venter is bisulcate-tricarenate, variably wide depending on the shape and size of the whorl. In some cases, the ventral clavi fade into a continuous ("Mercaticeras-like") lateral keel and both the two sulci and the ventral area appear narrowed; alternatively, the ventral clavi and the interspace between them drop into a broader lateral sulcus without forming an evident lateral keel, and the ventral area appears typically sunk.
Material and provenance. Two specimens from the Rosso Ammonitico section of Polino (Central Apennines). Specimen PO_LE10G2 (Figure 7a-d) is from the Lusitanicum subzone ( Figure 6); specimen PO_LEd1 (Figure 7e-h) is from an unknown bed.
Remarks. Specimen PO_LE10G2 is badly preserved and consists of a broken phragmocone (D = 26 mm) with a half-whorl body chamber eroded on both sides (Figure 7a,c). The fracture that cuts the terminal part of the body chamber has removed part of the internal phragmoconic whorls; the remaining is a 12 mm segment just preceding the onset of the body chamber (arrows in Figure 7a,c). The phragmoconic segment is better preserved than the body chamber and shows five closely spaced suture lines and three clavate ribs; the ventral area is poorly preserved, with the sulci and the central keel intensely eroded (Figure 7d).
Specimen PO_LEd1 (Figure 7e-h) is 18 mm in diameter, the whorl section is depressed (w/h = 8/6 = 1.3), and suture lines are not visible at eye or with eye lent, but they can be detected with an optical microscope, appearing very simplified (possibly a juvenile feature) and revealing that more than half of the preserved last whorl is part of the body chamber (arrows in Figure 7e,g). Ribs on the final whorl are 9 or 10, and some thickening (like thick striae) near the umbilicus is also observed in some of the interspace between ribs. The ventral clavi are evident and appear neatly cut along the break-off of the spine; the cut has left an oblique-flat surface that may mimic a forward projection of the clavi (Figure 7e). The ventral area is bisulcate-tricarenate.
Comparative description. The distinction between L. helenae and L. jeanneti is subtle and largely based on the narrower and less clavate ribs shown by the holotype of L. jeanneti (and also by the junior synonymous L. amuratica and L. attenuata; Figure 1e-g). In particular, the ribs in the final whorl of L. jeanneti show a periumbilical bump from which they extend straight or bend slightly backward and without significant thickening along the middle flank; an abrupt coarsening occurs on the ventral edge, producing a tubercle or clavi (Figure 1e,g). Also intercalated secondary ribs seem present in the holotype of the junior synonymous L. amuratica, which appear to fade out on the middle-inner flank, somewhat simulating their pairing with the preceding and more pronounced rib (Figure 1f).

Genus Neoleukadiella gen. nov.
Diagnosis. Neoleukadiella gen. nov. is characterized by overall evolute and quadratecompressed whorls with sinuous ribs ended by periumbilical and ventral tubercles, either simple or paired, or even welded at both ends by nodes to form fibulae. The ventral area is narrow and bisulcate-tricarenate, relatively wider in less compressed specimens. When present, paired ribs (merged by single or double nodes) irregularly alternate with simple ones, although the latter may be clearly prevalent or even the only present. Independently of being only simple or alternating with paired ribs, their number can be up to more than 20.
Remarks. Neoleukadiella gen. nov. is different from Leukadiella essentially in the ornamental pattern, the former being characterized by thinner and denser sinuous ribs (up to 20 per whorl or more), which are never clavate and are either simple or paired by periumbilical nodes, or even welded into fibulae when also ventral nodes are present. In Leukadiella, instead, clearly merging ribs and periumbilical tubercles are absent, and the thicker single ribs (rarely more than 10-12 per whorl) are ended by a ventral clavi.
The "giant" specimens from Canada described by Jakobs [25] are herein all interpreted as belonging to Neoleukadiella gen. nov., according to their ornamental pattern. This consists of coarse ribs that, however, differ from the typical ornamentation of Leukadiella because of their sinuous shape, the presence of an elongated umbilical tubercle instead of ventral clavi, and the frequent pairing into umbilical and ventral nodes.
Occurrence. Lower Toarcian, Bifrons Zone. Diagnosis. N. ionica comb. nov. is characterized by sinuous paired ribs, randomly alternating with simple ribs; the paired ribs may form fibulae and show an elongated umbilical node and a more rounded ventral node (Figure 1i). The simple ribs may become thinner and tend to lose tubercles in the mature stage (Figure 1i). The pattern of alternation between simple and paired ribs, as well as the occurrence of fibulae, appears overall random from the inner to the outer whorls, although simple ribs occur more frequently in the final whorl. In the holotype (herein refigured in Figure 1i) the whorl section is slightly compressed (w/h = 0.90) in the last whorl, at a diameter of 41 mm [50].
Comparative diagnosis. N. ionica comb. nov. and N. gallitellii comb. nov. (Pinna, 1965) [53] (by original designation, pp. 269-270, pl. 1, figures 1a−c, 3b) differ because, in the latter, the ribs are simple and never paired along the umbilical or ventral margin, though they as well depart from an elongated, forward-projected umbilical tubercle, and terminate in a well-marked, forward-projected ventral tubercle (Figure 1j; Figure 1k,l by synonymy). In the holotype of N. gallitellii comb. nov. (refigured in Figure 1j) the whorl section is quadrate (w/h = 1) in the last whorl, at a diameter of 38 mm [53].

Taxonomic Review and Phylogenetic Implications
The comparative analysis of morphological characters traditionally used to define Leukadiella species suggests that different rib patterns and shell structure are mutually associated and exclusive of some species only. On this basis, two distinct clusters of taxonomic characters can be defined, herein used to separate two morpho-groups: the Helenae Group and the Ionica Group. The marked diversity of these two groups could be bridged by the supposed (though not obvious) transition from the amuratica-attenuata-jeanneti ornamental pattern (herein considered as representing one species, namely L. jeanneti) to that of species of the Ionica Group ( Figure 10); alternatively, the lack of any phylogenetic link between the two groups would point towards their separate origin. In any case, the Helenae and Ionica groups are interpreted as representing two distinct genera. The name Leukadiella is maintained for the Helenae Group, whereas the new name Neoleukadiella gen. nov. is proposed for the Ionica Group; each genus includes two of the four species herein maintained valid: L. helenae, L. jeanneti, N. ionica comb. nov. and N. gallitellii comb. nov.
This revised taxonomy avoids the problem of accounting for a remarkable peculiarity in the systematics of hildoceratids: the inclusion in the same genus of species with straightclavate simple ribs, or thin-sinuous simple ribs with or without umbilical and ventral tubercles; or even paired ribs with umbilical nodes, or both umbilical and ventral nodes forming fibulae. Indeed, the differences in ornamental and structural patterns used to separate Leukadiella and Neoleukadiella gen. nov. appear consistent when compared with differences among other genera within hildoceratids or other groups; some examples are provided by Hildaites and Orthildaites; Mercaticeras, Pseudomercaticeras Merla, 1933 [85] and Merlaites Gabilly, 1974 [86]; Phymatoceras Hyatt, 1867 [14] and Chartronia Buckman, 1898 [87]; Paroniceras and Frechiella; and even Renziceras and Leukadiella.
The problem of the founder species of the Leukadiellinae is an unsolved question. H. propeserpentinus (Figure 3a) is frequently considered the most likely direct progenitor of L. ionica [25], whereas some coarse-ribbed Hildaites (e.g., H. forte Buckman, 1921 [19]) or even Orthildaites (Figure 3b) are usually assumed as ancestral to L. helenae, either directly or via Renziceras [17,[22][23][24]. Only the genus herein referred to as Leukadiella appears closely related and in morphological continuity with Renziceras, from which it can be derived by heterochronic truncation of the adult stage, characterised by evolute coiling and "Arieticeras-like" ornamentation (Figures 1a and 5d). Conversely, there is limited evidence for extending heterochrony also to the arising of N. ionica comb. nov. from Renziceras or even from L. amuratica (as proposed by Wendt [23]) or L. helenae (as proposed by Macchioni and Venturi [17]). As for the ancestor of the Renziceras-Leukadiella lineage, the basic morphological features displayed by Cingolites (Figure 5a-c) represent a liable prototype from which the Renziceras morphological pattern can be derived. In particular, the clavate ornamentation and the broad-coronate venter of Cingolites may bridge the Renziceras-Leukadiella stock with the strong-ornamented Hildaites or Orthildaites.
The ornamental pattern displayed by Cingolites is similar to the hildaitic coarse-ribbed pattern to a point that Howarth [65] considers this genus as synonymous of Hildaites. However, the distinguishing structural and ornamental features described by Sassaroli and Venturi [26] are noteworthy and justify a taxonomic distinction between Cingolites and Hildaites. In particular, the main features by which Cingolites differs from the coarse-ribbed Hildaites and Orthildaites (i.e., overall clavate ribs and broad coronate venter with wide sulci) are the same by which it appears comparable with Renziceras, and thus close to the Leukadiellinae. Despite this evidence, Sassaroli and Venturi [26] accommodate Cingolites in the subfamily Hildoceratinae. Furthermore, because both Renziceras and Cingolites have been recurrently acknowledged as morphologically and phylogenetically comparable either with Hildaites and Mercaticeras, it is surprising that Sassaroli and Venturi [26] avoided outlining any phylogenetic affinity between their new genus Cingolites and Renziceras.
The filling of the gap between Renziceras-Leukadiella and the Hildaites stock significantly attenuates the morphological peculiarity by which the arising of Leukadiella has been historically connoted. In addition, the ornamental pattern and the broad sulcatedtricarenated ventral area of Cingolites also suggest a possible relationship between this genus and Mercaticeras. In this view, the affinity remarked by diverse authors between Mercaticeras and Leukadiella reflects the sharing of an origin from a common group, rather than an ancestor-descendant relationship.
Tracing the origin of Neoleukadiella gen. nov. is instead more difficult, and available data point towards the two possible hypotheses outlined above: one based on the direct derivation from the Helenae Group [17,23] (see Figure 4) and one considering a distinct origin from an earlier hildaitic ancestor ( Figure 11). The latter hypothesis relies essentially on the assumption that the paired ribs link N. ionica comb. nov. directly to H. propeserpentinus [22,25]. However, the double nodes forming fibulae along the ribs of N. ionica comb. nov. are markedly different from the nodes and ribbing pattern of H. propeserpentinus (as well as from that of any other Hildaites species). This observation is not meant to exclude Hildaites as a possible ancestor but rather to remark that other alternatives can be considered, including that of a simple ribbed descent, either within the Hildaites group or a different one. It is worth noting that nodes and rib-pairing are features that may arise in species stemming from a simple-ribbed morphotype; as, for instance, in the case of Pseudomercaticeras, Phymatoceras and Merlaites, all with random nodes and coupled ribs though likely descending from the simple-ribbed Mercaticeras. As an implication of this reasoning, other alternatives deserve consideration, including that of a less coarsely ornamented variant of Leukadiella as a possible precursor of the Neoleukadiella (i.e., the Neoleukadiella 2 hypothesis in Figure 11); indeed, L. jeanneti with its broader ornamental variability, as herein interpreted (Figure 10), may likely be this precursor.

Internal Clocks and "Unconventional" Evolutionary Mechanisms
The origin of the Leukadiellinae seems consistent with the renowned role of Hildaites as a "prolific" founder of early-middle Toarcian groups [17,26,55,77]; a question therefore arises on the reasons why Hildaites becomes prolific. Furthermore, remains unclear how paired ribs with nodes and fibulae arise abruptly, manifesting together with simple ribs and in an overall irregular and inconstant pattern. Based on the above results, heterochrony and anagenetic trends only partly account for the diversification of the Leukadiellinae from the Hildoceratinae morphotype. Therefore, other "unconventional" mechanisms may have determined the stemming of new phenotypic variants. The speculation herein put forward is that both the prolificity of the Hildaites stock and the abrupt arising of "peculiar" ornamental patterns can be put in relation with some kind of "morphogenetic susceptibility" triggered by environmental perturbation [44,[88][89][90][91]. In this view, the abrupt arising of morphological novelties may reflect alteration, under unfavourable environmental conditions, of "internal clocks" controlling developmental processes or even silent gene re-activation mechanisms [48].

Internal Clocks and "Unconventional" Evolutionary Mechanisms
The origin of the Leukadiellinae seems consistent with the renowned role of Hildaite as a "prolific" founder of early-middle Toarcian groups [17,26,55,77]; a question therefor arises on the reasons why Hildaites becomes prolific. Furthermore, remains unclear how paired ribs with nodes and fibulae arise abruptly, manifesting together with simple rib and in an overall irregular and inconstant pattern. Based on the above results, hetero chrony and anagenetic trends only partly account for the diversification of the Leukadiel linae from the Hildoceratinae morphotype. Therefore, other "unconventional" mecha nisms may have determined the stemming of new phenotypic variants. The speculation herein put forward is that both the prolificity of the Hildaites stock and the abrupt arising of "peculiar" ornamental patterns can be put in relation with some kind of "morphoge netic susceptibility" triggered by environmental perturbation [44,[88][89][90][91]. In this view, th abrupt arising of morphological novelties may reflect alteration, under unfavourable en vironmental conditions, of "internal clocks" controlling developmental processes or even silent gene re-activation mechanisms [48].
One such mechanism is described by Guex [44] based on cases of ammonite lineage which, concomitantly with a deterioration of environmental conditions, tend towards simplification of their morphology by recalling ancestral patterns (i.e., proteromorphosis) The reappearance of ancestral ornamental features introduces new (though "recycled" phenotypic patterns at the end of a lineage, mimicking a reiterative evolutionary trend starting and ending with comparably simplified morphotypes (i.e., homeomorphism) The ending homeomorph with its simplified morphology often preludes to the arising o a new lineage or even to an evolutionary radiation [44]. This phenomenon has been cor related with the effect of environmental stress and referred to as "reinitialization of evo lutionary clocks" [44] (hereinafter REC).
Whatever the mechanism by which the archaic genetic information is maintained and reactivated, the frequently observed re-occurrence of archaic morphologies, also in patho One such mechanism is described by Guex [44] based on cases of ammonite lineages which, concomitantly with a deterioration of environmental conditions, tend towards a simplification of their morphology by recalling ancestral patterns (i.e., proteromorphosis). The reappearance of ancestral ornamental features introduces new (though "recycled") phenotypic patterns at the end of a lineage, mimicking a reiterative evolutionary trend starting and ending with comparably simplified morphotypes (i.e., homeomorphism). The ending homeomorph with its simplified morphology often preludes to the arising of a new lineage or even to an evolutionary radiation [44]. This phenomenon has been correlated with the effect of environmental stress and referred to as "reinitialization of evolutionary clocks" [44] (hereinafter REC).
Whatever the mechanism by which the archaic genetic information is maintained and reactivated, the frequently observed re-occurrence of archaic morphologies, also in pathological or injured ammonites [92], is an indication that genetic information relative to ancestral structures is not (completely) lost during phylogenesis [93]. In addition, experimental evidence indicates that environmental stress (thermal or chemical) during the early developmental stages of organisms may alter the ontogenetic expression of phenotypes [88,94,95]; these developmental effects may be key to the understanding of how archaic morphologies are preserved and resumed. In this view, it is likely that, although different, heterochrony and REC may share a mechanism based on the susceptibility of internal (developmental and evolutionary) clocks to external triggering causes (altered/distressed environmental conditions); such mechanisms may be more varied than heterochrony and REC, and manifest diversely in the fossil record.
Differently from REC processes, the relevance of heterochrony in fossil-documented evolutionary events has been poorly considered in the light of its affinity with environmental triggering. Nevertheless, these processes have relevant implications for evolutionary mechanisms because the onset and persisting of ecologically induced phenotypic/developmental changes can be conceived as an epigenetic-like mechanism. In a similar scenario, some interesting implications disclose by assuming that, at least in some cases, the lethal consequences of perturbed developmental processes may be attenuated by the emerging, via REC-like processes, of a "coherent phenotypic anomaly". Coherent phenotypic anomaly is referred to any "abnormal" feature that reproduces a morphological pattern normally displayed by other (usually ancestral) groups. It is thus possible that these stress-induced morpho-genetic changes determine tolerable (though abnormal) phenotypic variants within populations; these are subject to Darwinian selection as any other variant, resulting in their elimination or diffusion. Once screened by selective processes, the uprising phenotypic pattern would be maintained throughout the recession of the external pressure, leading to the ensuing of a morphologically diversified population.
Because environmental stress also results in the overall restriction of favourable living conditions, the size of the ensuing modified populations is influenced and subject to restraining in terms of number of individuals and geographical distribution. At their stemming, the new morphotypes may have a low frequency, ranging between that of pathological individuals and that of small populations struggling to persist a bottle necklike phase. In this view, the interplay of environmental stress and developmental processes can be considered as an unconventional morphogenetic mechanism that, in addition to classical mutation events (as well as frequently conducive to lethal consequences), prompts Darwinian selection by expanding phenotypic variability. Moreover, because developmental alteration may cause apparently abrupt morphological changes, marked by the reoccurrence of "pre-evolved" structures, phenotypic variation is broadened in a non-gradualist way. Indeed, "monstrous" phenotypes (in the form of coherent phenotypic anomaly) that persist through the stressful conditions by which they spur provide material for selective pressure to burst speciation and marked (macro?) diversification.

Pathological Morphotypes and "Coherent Phenotypic Anomaly"
The above scenario, although speculative, well describes phenomena recorded in the fossil record and interpreted within the frame of heterochrony, atavism, and REC [28,44]. In addition, pathological specimens documented in the literature show, apart from cripple-like anomalies affecting ornamental features [92], also abnormal morphology of the kind herein referred to as coherent phenotypic anomaly. A typical case is that in which the two sides of an ammonite are different, one reproducing the exact pattern of a related species [93]. However, the limits between "cripple" and "coherent" morphological anomalies are not always obvious. In this respect, an interesting example is provided by two abnormal specimens of Hildoceras, one figured by Renz [49] (pl.7, figure 4) and one by Buckman [96] (p. DCCLXXIIIA, figures 1-4). Buckman's "dysmorph" is a specimen of Hildoceras bifrons Bruguière, 1789 [97] displaying on one side an abnormal morphology perfectly coherent with the ornamental pattern of its ancestor Hildoceras laticosta Bellini, 1900 [98] (Figure 12a,b); Renz's "pathological form" shows ribs on the last whorl that tend to merge along the ventral margin, by this recalling the ornamental pattern typical of N. ionica comb. nov. (Figure 12c). Although different in many respects from those developed by Neoleukadiella gen. nov., the ventral "pseudonodes" of this abnormal Hildoceras provide evidence for the possible occurrence of such morphological features as a consequence of some kind of dysfunction.
The assumption herein made is that the abrupt arising of peculiar ornamental patterns, such as those displayed by the Leukadiellinae (in particular by Neoleukadiella gen. nov.), can be foreseen as one among other cases of stress-induced altered development and/or re-occurrence of ancestral-like morphological features, rather than representing inheritance from a direct ancestor retaining similar characters (i.e., Hildaites species or others retaining a precursor homology in their ornamental pattern). Coarse ribs with tubercles, nodes and spines are indeed frequently displayed in diverse ammonite groups, some of which within lineages ancestral to the Hildoceratinae (as for instance Bouleiceras, among others). The assumption herein made is that the abrupt arising of peculiar ornamental patterns, such as those displayed by the Leukadiellinae (in particular by Neoleukadiella gen. nov.), can be foreseen as one among other cases of stress-induced altered development and/or re-occurrence of ancestral-like morphological features, rather than representing inheritance from a direct ancestor retaining similar characters (i.e., Hildaites species or others retaining a precursor homology in their ornamental pattern). Coarse ribs with tubercles, nodes and spines are indeed frequently displayed in diverse ammonite groups, some of which within lineages ancestral to the Hildoceratinae (as for instance Bouleiceras, among others).
Although it is not possible to unambiguously link any of such mechanisms to the emerging of the specific morphological changes displayed in the fossil record by species of the Leukadiellinae, they are a valid alternative to supposed (though stratigraphically not evident) continuous and gradual anagenetic trends originating from species of Bouleiceras or Hildaites. Moreover, their possible relationship with environmental stress accounts for the reduced areal and stratigraphic distribution of the ensuing species, as well as for the poor number of findings and small size of individuals, which are all features that one would expect from populations that are struggling to survive under unfavourable conditions. The migration to different (and possibly more favourable) environments may account for the increasing of average growth size, as in the case of the North American representatives (herein interpreted as representatives of Neoleukadiella gen. nov.).

Environmental Stress, Species Resilience and Biodiversity Recovery
One of the most invoked causes of protracted environmental perturbation with evolutionary implications is the giant volcanism and the resulting effects on water/atmosphere chemistry and global climate [27,38,39]. Some major extinctions appear correlated with these events and also with regional-global marine regressions, in some cases associated with negative trends in δ 13 C [44]. Among these there is the late Pliensbachian-early Toarcian crisis, triggered by the Karoo magmatic event and recorded by marine transgression-regression events and black shale deposition [40,41,43,45]. The spreading of the Hildoceratinae is commonly interpreted in terms of an evolutionary radiation following this environmental crisis and the subsequent extinctions [42,44,99].
According to the available biostratigraphic data, the Leukadiella-Neoleukadiella stock occurs in the Bifrons Zone (Lusitanicum Subzone at the Polino section; Figure 6). This interval post-dates the early Toarcian crisis and recovery of the Hildoceratinae, when Cingolites and Renziceras (or even an unknown direct ancestor of Neoleukadiella gen. nov.) are supposed to appear. Notwithstanding the uncertainties on the origin of Neoleukadiella gen. nov., even in the case of the origin of Leukadiella there is a noticeable stratigraphic gap Although it is not possible to unambiguously link any of such mechanisms to the emerging of the specific morphological changes displayed in the fossil record by species of the Leukadiellinae, they are a valid alternative to supposed (though stratigraphically not evident) continuous and gradual anagenetic trends originating from species of Bouleiceras or Hildaites. Moreover, their possible relationship with environmental stress accounts for the reduced areal and stratigraphic distribution of the ensuing species, as well as for the poor number of findings and small size of individuals, which are all features that one would expect from populations that are struggling to survive under unfavourable conditions. The migration to different (and possibly more favourable) environments may account for the increasing of average growth size, as in the case of the North American representatives (herein interpreted as representatives of Neoleukadiella gen. nov.).

Environmental Stress, Species Resilience and Biodiversity Recovery
One of the most invoked causes of protracted environmental perturbation with evolutionary implications is the giant volcanism and the resulting effects on water/atmosphere chemistry and global climate [27,38,39]. Some major extinctions appear correlated with these events and also with regional-global marine regressions, in some cases associated with negative trends in δ 13 C [44]. Among these there is the late Pliensbachian-early Toarcian crisis, triggered by the Karoo magmatic event and recorded by marine transgression-regression events and black shale deposition [40,41,43,45]. The spreading of the Hildoceratinae is commonly interpreted in terms of an evolutionary radiation following this environmental crisis and the subsequent extinctions [42,44,99].
According to the available biostratigraphic data, the Leukadiella-Neoleukadiella stock occurs in the Bifrons Zone (Lusitanicum Subzone at the Polino section; Figure 6). This interval post-dates the early Toarcian crisis and recovery of the Hildoceratinae, when Cingolites and Renziceras (or even an unknown direct ancestor of Neoleukadiella gen. nov.) are supposed to appear. Notwithstanding the uncertainties on the origin of Neoleukadiella gen. nov., even in the case of the origin of Leukadiella there is a noticeable stratigraphic gap separating it from Renziceras. This stratigraphic gap may be a consequence of the amplification of the low frequency of Leukadiella in the lower Toarcian, an interval where also other groups become infrequent [24]. In the Apennines, this interval corresponds to a sterile or poorly fossiliferous part of the Rosso Ammonitico Formation (Figure 6), known as "unnamed zone" because of the lack of biostratigraphic markers [1,100]; sediments of this interval are often affected by anoxic events [55].
In this view, the restriction to the Bifrons Zone of the occurrence of Leukadiella may reflect a biased first appearance datum; indeed, one finding of Leukadiella below the Bifrons Zone is reported in the literature [17]. This may indicate that earlier populations of Leukadiella, if ever existed, were even smaller than during the Bifrons Zone, when they grew somewhat larger though never reaching the wealth of other coeval groups (e.g., Hildoceras and Mercaticeras, among others). Possibly, pulses of resurging environmental stress during the Bifrons Zone may have impacted on these striving populations of Leukadiella, determining their final extinction or even enhancing the arising of the abnormal and somewhat atavistic morphological features (e.g., thin-sinuous simple ribs and paired ribs with umbilical and ventral nodes) displayed by Neoleukadiella species (in accordance with the "Neoleukadiella 2" hypothesis; Figure 11).
More in general, the sterile or poorly fossiliferous intervals in the lower Toarcian Rosso Ammonitico may record a period of recovery and resilience of the few species that survived the late Pliensbachian-early Toarcian extinction events; locally, restrained populations gave rise to new morphotypes, only some of which persisted until the next extinction or the next evolutionary radiation. The evolutionary radiations following extinctions are traditionally correlated with the simplification of morphological patterns at the end of the surviving lineage [44]; this simplification is interpreted as a way for the survivals to regain "adaptation potential" and evolvability towards new complexity. However, it is also likely that the "peculiar" phenotypes emerging during these periods of environmental and biological crisis are as well prime to the stemming of morphologically new evolutionary lineages, by which biodiversity loss is partly healed (or even increased).
As an example, despite their overall simple morphology, the Hildaites species spreading immediately after the crisis display a number of different (sometimes contrasting) ornamental variants, as for instance: rib thinning and increasing in number or thickening and spacing out; trends towards more sinuous or strait ribs; umbilical tubercles and nodes grouping ribs or fading of the umbilical termination of ribs; this latter trend leading to the complete disappearance of the umbilical segment and the forming of the umbilical wall and smooth band preluding to the Hildoceras morphotype [101,102]. Following this early Toarcian radiation, while the Cingolites-Renziceras-Leukadiella-Neoleukadiella branch was not particularly successful, most of the ornamental and structural novelties introduced (or re-introduced) by the early Hildoceratinae persisted in other groups such as Mercaticeras, Pseudomercaticeras, Merlaites and Phymatoceras, significantly contributing to the middle Toarcian differentiation of diverse ammonite lineages.

Conclusions
The rarity of specimens of Leukadiella (only 70 findings worldwide documented in the literature) is a renowned obstacle to their understanding and classification, an issue further complicated by the peculiar morphology of this taxon compared to that of other Hildoceratidae. It is not simple to define distinctive morphological traits for tracing the limits and defining the relationships between most of the historically established species, or for reconstructing their phylogeny with respect to other groups. Moreover, the supposedly related genera as well are infrequent in the stratigraphic record. Renziceras, for instance, is even less common than Leukadiella, with only seven findings in the Apennines and a total of nine documented worldwide.
The comparative analysis of taxonomic characters (largely based on figures) indicates that the morphological differences that authors have referred to for distinguishing up to eleven species smooth to a point that their number reduces to only four; these are parted in the two genera Leukadiella and Neoleukadiella gen. nov. as follows: L. helenae, L. jeanneti, N. ionica comb. nov. and N. gallitellii comb. nov. Only genus Leukadiella appears closely related and in morphological continuity with Renziceras; the origin of Neoleukadiella gen. nov. remains dubious and may be tentatively derived either from Leukadiella or from an unknown early Toarcian progenitor. Both the Renziceras-Leukadiella branch or the putative unknown progenitor of an independent Neoleukadiella gen. nov. likely descend from the Hildaites stock. In the case of Renziceras-Leukadiella, an intermediate morphotype has been identified in Cingolites, an Apennine genus morphologically bridging the coarse-ornamented Hildaites and the clavate-ribbed Renziceras. In addition, some ornamental and structural features of Cingolites also suggest a possible relationship with Meracaticeras (likely the progenitor of other species with paired ribs and nodes, such as Pseudomercaticeras, Merlaites and Phymatoceras).
The simple ribs with double tubercles of N. gallitellii comb. nov. and, even more, the fibulate ribs of N. ionica comb. nov., represent brand-new ornamental patterns within the Hildoceratidae that cannot be directly derived from any pre-existing ornamental pattern within Hildaites, Orthildaites, Renziceras or other lower Toarcian Hildoceratinae; therefore, linking Neoleukadiella with other taxa requires a morphological jump. In the light of this unavoidable discontinuity, also a parental relationship with Leukadiella deserves consideration. For instance, Leukadiella and Neoleukadiella gen. nov. may be bridged based on some "peculiar" features by which L. jeanneti is distinguished with respect to L. helenae; these may include the tendency towards less thick ribs and fading of the umbilical segment of secondary ribs, altogether resembling (and perhaps being precursor of) ribs paired by umbilical nodes.
As a conceptual exercise, the origin of the straight-clavate ribs in Cingolites (later accentuating in Renziceras-Leukadiella) and the arising of the sigmoidal-fibulate ribs in Neoleukadiella gen. nov. may be interpreted in the frame of developmental distressing, instead of a direct inheritance via anagenetic trends from an ancestor already possessing similar characters. Altered development may result in abnormal (i.e., pathological) individuals or in phenotypic effects comparable to those emerging during heterochrony and REC-like processes. These include the reactivation of silenced developmental schemes by which ancestral structures are resumed. The resumed characters may retain a "coherent morphology" by which they may not result lethal, although being the expression of a pathological or disease-like induction effect. The ensuing modification may be conceived as any other mutation-driven morphological change, and thus reveal neutral, advantageous or disadvantageous under selective pressure. Therefore, the unconventional evolutionary processes herein invoked are not meant as in contrast with Darwinian selection, but rather as a morphogenetic mechanism complementary to punctual gene mutation, potentially capable of expanding in a non-gradualist way the phenotypic variability on which selective processes operate.
From an evolutionary perspective, the problem arising concerns the mechanism by which ecologically induced developmental (regulatory) changes become "genetically fixed" in order to persist on the long-term and give rise to a new phyletic lineage, rather than resulting in an abnormal phenotype that would be wiped out within one generation or in the next few ones. The existence of a similar mechanism has implications as to the meaning and impact of epigenetic-like processes during evolution. It is worth noticing that the abrupt emerging of new morphologies under the control of this overall epigeneticlike mechanism is not viewed as a direct adaptive response to environmental pressure, but rather as a consequence of a stress-induced dysfunction that casually reintroduces pre-existing morphological patterns. As a paradox, the main impact of epigenetic-like processes in determining marked morphological modifications during evolutionary events would result from the resuming of "preformed" structures. While the above hypotheses are evidently speculative, they are conceptually intriguing in the light of exploring evolutionary pathways from a paleontological perspective, specifically by relying on the reading of the "Leukadiella event" as it appears from the stratigraphic record.