Introduction

The evolution of life has never been a straight line, particularly not for ammonoids. As documented by many authors, the ammonite conch underwent several morphological changes during ontogeny, possibly reflecting changes in habitat and habit (e.g., Drushchits et al. 1977; House 1996; Westermann 1996; Klug 2001; Lukeneder et al. 2010; Ritterbush et al. 2014; Robin et al. 2014; Lukeneder 2015; Mironenko and Rogov 2015; Naglik et al. 2015a). These ontogenetic changes can be profound to subtle. Examining conchs of adult ammonoids bears the advantages for the studying researchers that they commonly formed distinct adult modifications such as constrictions or changes in coiling (e.g., Davis et al. 1996; Klug et al. 2015b) and that intraspecific variability appears to be lower (De Baets et al. 2015), thus facilitating species determination. By contrast, determination of juvenile forms, which are often more common, is sometimes difficult due to the lack of adult characters and because they display a different kind of variability (e.g., higher degree of interspecific overlap, dimorphism is not or poorly expressed; e.g., Davis et al. 1996; De Baets et al. 2013, 2015; Klug et al. 2015a). In any case, knowledge of the whole ontogeny is important to understand ammonoid systematics, phylogeny, and ecology (see De Baets et al. 2013; Klug et al. 2015a).

Ammonoid palaeoecology is difficult to assess, because direct observation of ammonoid behaviour is impossible and we have to rely on inferences that can be drawn from conch and jaw morphology. In this context, Tendler et al. (2015) have recently explored the distribution of shell shapes in morphospace. Based on the topography of this morphospace, they evaluated possible links between conch morphology and the optimisation for certain tasks. These tasks are (i) economy of shell material (see also Raup 1967), (ii) optimal hydrodynamics (Chamberlain 1976; Jacobs 1992; Naglik et al. 2015b), and (iii) rapid shell growth (= high whorl expansion; compare Bucher et al. 1996). However, Tendler et al. (2015) found that the morphospace of adult ammonoid conchs filled a pyramid with five vertices. Accordingly, they interpreted the remaining two vertices as being occupied by shell morphologies optimized for combinations of two tasks, namely compactness with economy of shell material (iv) and compactness with hydrodynamics (v).

Fig. 1.

Geological map of the eastern Anti-Atlas (modified after Korn et al. 2005).

In this paper, we want to explore these links using the well preserved Carboniferous materials from Morocco. After some quiescence following the pioneering works on Early Carboniferous (Mississippian) ammonoids from northern Africa (e.g., Delépine 1941; Pareyn 1961; Hollard 1958), research on these fossils is currently experiencing a renaissance (Korn et al. 1999; 2002, 2003a, b, 2005, 2007, 2010, 2014; Ebbighausen et al. 2004, 2010; Bockwinkel and Ebbighausen 2006; Klug et al. 2006; Bockwinkel et al. 2010). Possibly due to the quality of the outcrop (large plains covered in loose gravel with fossils aligned according to their source rock), several Carboniferous ammonoid assemblages (including the Entogonites-bearing assemblage described by Korn et al. 2005; see map in Fig. 1) had been overlooked in the eastern Anti-Atlas (Morocco) although they are often well preserved, sometimes rather highly diverse and the contained forms display an impressive disparity. These assemblages yielded both juvenile and adult conchs of several, morphologically quite disparate species. This morphological disparity includes not only differences between different taxa but also within taxa between different growth stages (see also Klug 2001).

In the case of the nodules with Entogonites, the fossil-bearing rocks are scarce and not so easy to find. The nodules containing the Entogonites assemblage are usually flat, brownish and contain fossils in nearly rock forming numbers. Some of the rocks may actually represent gutter casts, which is supported by the fact that some of the nodules contain only juvenile specimens and that the nodules have a flat cross section and appear to be elongate. It is well visible that the contained ammonoids are reasonably diverse because their morphology differs strongly. The Entogonites assemblage contains all of the main types of “monomorphic” (planispirally coiled, bilaterally symmetric) ammonoids including oxyconic (adult Calygirtyoceras), serpenticonic (Nomismoceras), cadiconic (juvenile Calygirtyoceras), quadrate (juvenile Entogonites), and globose (adult Goniatites) to extremely spindle-shaped forms (juvenile Goniatites). A closer look revealed that certain morphologies are absent in the juveniles and others in the adults of this assemblage. This, in turn, led us to question whether this absence of certain conch shapes could be caused by sorting or other taphonomic processes or whether this is a primary signal, where juveniles or adults, respectively, simply did not produce conchs with such shapes.

Accordingly, the aims of our study are to: (i) describe the morphology of the members of the Entogonites assemblage, (ii) document, which morphologies are adult-only or juvenile-only, and (iii) discuss potential explanations for the patterns.

Institutional abbreviations.—GPIT, Institut für Geowissenschaften, Eberhard Karls Universität Tübingen, Germany; MB.C, Museum für Naturkunde, Berlin, Germany; PIMUZ, Paläontologisches Institut und Museum, Zürich University, Switzerland.

Other abbreviations.—ah, apertural height; dm, diameter; uw, umbilical width; WER, whorl expansion rate; ww, whorl width (measured between ribs). Conch parameters after Raup (1966: 1294; see below for the equations to determine these values): D, the distance between the generating curve and the coiling axis; S, the shape of the generating curve (“equivalent to the cross-sectional shape of the tube”); W, whorl expansion rate (“the rate of increase in the size of the generating curve per revolution”).

Material and methods

The recently collected, well preserved specimens were all found in the pebble-covered reg-like desert (reg: gravel-covered flat area) about 12 km south of the oasis of Dar Kaoua. This assemblage described herein was partially presented by Korn et al. (1999, 2005, 2007) including one of the stratigraphically oldest occurrences of a member of the genus Goniatites (G. lazarus) and a new species of Entogonites (E. saharensis). All material was collected in a plain, about 10 to 12 km south of the Kasbah (fortified Moroccan building) of Dar Kaoua (Tafilalt, SE of Erfoud, Morocco; Fig. 1). The reg-like plain is largely flat with shallow hills covered by scree and very little true exposure. Nevertheless, some fossiliferous layers locally yield well-preserved ammonoids, orthocerids, nautilids, gastropods, trilobites, bryozoans, brachiopods, and crinoids. These layers can be vaguely traced by following the fossil occurrences. The relative abundances of ammonoids are listed in Table 1.

As far as measurements and ratios are concerned, we followed the suggestions published by Korn (2010) as well as Klug et al. (2015a). Also, we use technical terms in the sense used in these two papers. To calculate the classical Raupian parameters W, S, and D, we used the original equations by Raup (1966, 1967). Tendler et al. (2015) repeated the equations needed to determine the values:

where a is the radius from the coiling axis to the venter, b is the same value of the preceding whorl, x is the umbilical width measured as a ratio from the coiling axis, y is the whorl width and z is the whorl height.

Systematic palaeontology

Table 1.

Quantity of specimens of different taxa in the new material from the early late Viséan, 12 km south of the oasis of Dar Kaoua.

Table 2.

Measurements (in mm) and ratios of Goniatites lazarus. Abbreviations: ah, apertural height; D, the distance between the generating curve and the coiling axis; dm, diameter; S, the shape of the generating curve; uw, umbilical width; W, whorl expansion rate; WER, whorl expansion rate; ww, whorl width (measured between ribs).

Fig. 2.

Juvenile (neanic) specimens of Goniatites lazarus Korn, Klug, and Mapes, 2005, Entogonites saharensis Korn, Klug, and Mapes, 2005, and Entogonites bucheri sp. nov. All from the early late Viséan, 12 km SE of Dar Kaoua (Tafilalt, Morocco). A. PIMUZ 31512, two neanic G. lazarus specimens (white arrows) in ventral views (A₁), note the associated E. saharensis and orthocones. Enlarged G. lazarus in lateral view (A₂); the narrow umbilicus and the umbilical ridge (A₃). B. PIMUZ 31512, a juvenile G. lazarus, a small juvenile and a subadult E. saharensis; G. lazarus in ventral (B₁) and lateral (B₂) views. C. PIMUZ 31520, detail of a fully grown Maxigoniatites saourensis (Pareyn, 1961) with G. lazarus, 3 juvenile E. saharensis, and a subadult E. saharensis and a hatchling of E. bucheri sp. nov. (arrowed); C₂, detail of C₁, showing G. lazarus and E. bucheri sp. nov. All specimens whitened with NH4Cl-sublimate. Scale bars 10 mm.

Fig. 3.

Adult specimen of Goniatites lazarus Korn, Klug, and Mapes, 2005 (PIMUZ 31514) from the early late Viséan, 12 km SE of Dar Kaoua (Tafilalt, Morocco), in dorsal (A), lateral (B), and ventral (C) views. All specimens whitened with NH₄Cl-sublimate

Fig. 4.

Comparison of cross sections, measurements, and ratios of Goniatites lazarus and other ammonoids. A. Goniatites lazarus, subadult (MB.C. 25130, A₁) and juvenile (PIMUZ 31510, A₂) from early late Viséan, 12 km SE of Dar Kaoua (Tafilalt, Morocco). B. Goniatites fimbriatus, MB.C.13299 (after Korn et al. 2008: fig. 23A) from Nehden, Viséan, Rhenish Mountains, Germany. C. Juvenile Kornia citrus, MB.C.10202.1 (after Ebbighausen and Bockwinkel 2007: fig. 30B) from early Tournaisian, Aguelmous (Tafilalt, Morocco); note the conch shape and umbilical ridge, which is similar to juvenile G. lazarus. Whorl width (ww) and umbilical width (uw) indexes (D) and whorl expansion rate (WER) (E) of G. lazarus, G. fimbriatus, and K. citrus.

Table 3.

Measurements (in mm) and ratios of Calygirtyoceras darkaouaense. Abbreviations: ah, apertural height; D, the distance between the generating curve and the coiling axis; dm, diameter; S, the shape of the generating curve; uw, umbilical width; W, whorl expansion rate; WER, whorl expansion rate; ww, whorl width (measured between ribs).

Fig. 5.

Ammonoids from the early late Viséan, 12 km SE of Dar Kaoua (Tafilalt, Morocco). A, C. Calygirtyoceras darkaouaense Korn, Klug, and Mapes, 1999. A. PIMUZ 31516 (A₁), Goniatites lazarus with a juvenile specimen in oblique-ventral (A₂), lateral (A₃), and ventral (A₄) views. C. PIMUZ 31513, fragmentary adult in lateral view (C₁), cross section (C₂). B. Entogonites bucheri sp. nov., PIMUZ 31509, holotype in lateral (B₁) and ventral (B₂) views. All specimens whitened with NH₄Cl-sublimate.

Fig. 6.

Mass occurrence of juvenile ammonoids from the early late Viséan, 12 km SE of Dar Kaoua (Tafilalt, Morocco). A. PIMUZ 31518 with 4 specimens of Nomismoceras sp., 3 Prolecanites sp., 4 Entogonites saharensis Korn, Klug, and Mapes, 2005, 1 Calygirtyoceras darkaouaense Korn, Klug, and Mapes, 1999, and 2 Bollandites sp. B. PIMUZ 31508, overview; many taxa are indicated in the figure. All specimens whitened with NH₄Cl- sublimate.

Table 4.

Measurements (in mm) and ratios of Entogonites saharensis and E. bucheri sp. nov. Abbreviations: ah, apertural height; D, the distance between the generating curve and the coiling axis; dm, diameter; S, the shape of the generating curve; uw, umbilical width; W, whorl expansion rate; WER, whorl expansion rate; ww, whorl width (measured between ribs).

Fig. 7.

Morphospace (UWI, umbilical width index; WER, whorl expansion rate; CWI, conch width index) with data of 3015 species of Palaeozoic ammonoids with additional points of the ammonoids from the early late Viséan of the Tafilalt (Morocco). Note that data points of juveniles (empty symbols) of Calygirtyoceras darkaouaense and Goniatites lazarus lie at the edge or outside of the cloud of 3015 species. Graph produced using JMP 11.

Discussion

As far as the early late Viséan ammonoid fauna from the Tafilalt is concerned, the extreme morphology of some juvenile shells in combination with the often fundamentally different adult morphology is striking. One might suggest that these differences are rooted in the widely documented juvenile (post-hatching) morphological intraspecific variability of mollusk shells (e.g., Jacobs 1992; Jacobs et al. 1994; Jacobs and Chamberlain 1996; Urdy et al. 2010a, b; De Baets et al. 2013, 2015; Naglik et al. 2015a). This is in contrast with the fact that the conch morphology of some of the juveniles of early late Viséan ammonoids lies at the edge or even outside of the morphospace of adult Palaeozoic ammonoids (Figs. 7, 8). To put it differently: some juvenile morphologies are inexistent in adult ammonoids (at least in the Entogonites assemblage; we are not aware of adult ammonoids with a conch morphology remotely similar to juvenile Goniatites lazarus). The morphologies (see also the illustrations in Fig. 8; the terminology of Korn 2010 is given in brackets where it differs from the classical terms) we found in the major growth stages of the Entogonites assemblage are:

This leads to the question of whether this ontogenetic distribution of shell morphologies roots purely in morphogenetic constraints or whether there are also some aspects that can explain shell morphology, which root in the selection for distinct shapes. There are limits as to how to test these hypotheses since direct observation of these extinct animals is impossible. We can still discuss this issue a bit further by, for example, employing the set of hypotheses published recently by Tendler et al. (2015). In their paper, they explored possible adaptations and trade-offs throughout ammonoid evolution. They followed other authors (e.g., Arnold 1983; Farnsworth and Niklas 1995; Alexander 1996) in the assumption that organisms have to be capable to deal with differing “tasks” that often lead to trade-offs in constructional morphology. Tendler et al. (2015) employed the theory of Pareto Optimality (named after the economist Vilfredo Pareto), which is based on the fields of economy and engineering (Sen 1993; Shoval et al. 2012; Tomoiagă et al. 2013). Accordingly, the morphology of most organism groups would fill a polygon or polyhedron in the morphospace. Data points of specimens that are positioned in a corner of the polygon or polyhedron would represent forms optimized to fulfil one task predominantly. Such forms were dubbed archetypes by Tendler et al. (2015). Consequently, if the shape of a body part of an animal lies between two or more archetype positions, we can assume a limited optimization for several tasks, linked with trade-offs with regard to the others.

Fig. 8.

Ammonoid morphospace after Tendler et al. (2015) using the Raupian parameters W (whorl expansion rate), S (shape of the generating curve), and D (distance between the generating curve and the coiling axis). Note that the juveniles (empty symbols) of Calygirtyoceras darkaouaense and Goniatites lazarus (in contrast to the respective adults) lie outside of the five-sided pyramid. Juvenile Nomismoceras sp. plots near vertex 1, Entogonites spp. between 1 and 4, and juvenile Bollandites sp. near 4. A Devonian pinacitid is depicted near vertex 3 to illustrate the conch morphology.

Fig. 9.

Reconstruction of the ammonoid fauna from the early late Viséan,12 km SE of Dar Kaoua (Tafilalt, Morocco). A. Illustration of the habitat. B. The main components of the ammonoid assemblage.

In their study, Tendler et al. (2015) applied these ideas to a dataset of mostly adult ammonoids from all major groups. They discovered that ammonoids filled a space in the Raupian W-S-D morphospace (Raup 1966, 1967) that has a shape of a pyramid with five vertices (Fig. 8). These vertices correspond with morphologies representing optimization for shell growth, hydrodynamics, economy of shell material, compactness + economy of shell material, and compactness + hydrodynamics (i.e., the last two represent combinations of two optimizations each). When applied to the early late Viséan ammonoids from Morocco, this would imply differing optimizations depending on ontogenetic stages:

The absence of certain optimizations (independent of their functional interpretations) in certain ontogenetic stages might imply that juvenile conch morphologies had to be optimised partially to other tasks than adults. Interestingly, no juvenile of the Entogonites assemblage lies close to the vertices representing optimization for hydrodynamics and shell growth. These results coincide with findings of Jacobs (1992), Jacobs et al. (1994), and Jacobs and Chamberlain (1996), according to which the maximum sustainable swimming velocity depends strongly on size and shell morphology (reviewed in Naglik et al. 2015b). Additionally, large ammonoids with compressed shells (e.g., oxycones) could have reached the highest swimming velocities, while at small conch sizes, ammonoids with depressed shells (cadicones and spherocones) could swim faster (although still not fast).

Possibly, the congruence of this conclusion of Jacobs (1992) and Jacobs and Chamberlain (1996) with the distribution of juvenile morphologies versus adult morphologies is a coincidence. Also, Tendler et al. (2015) interpreted the conch morphologies with depressed shells as being optimised for compactness in combination with economy of shell material or hydrodynamics. It thus appears that such a compact conch morphology was advantageous for juvenile ammonoids of the Viséan. Compactness of the conch morphology has effects on swimming velocities and endurance by reducing drag, it has effects on reproduction through the reduction of embryonic shell size (De Baets et al. 2012) and on predation as individuals with similar width have a smaller diameter and thus might be harder to detect and catch. The hypothesis that juvenile conchs were optimized for tasks different from those of adults (or for similar tasks accomplished by different morphological means) finds a further, however weak, support in the fact that some of the slabs from the early late Viséan of the Tafilalt are covered by hundreds of juveniles of various ammonoid species (Figs. 6, 9): on the one hand, these associations could represent taphocoenosis of juveniles that shared habitats different from those of adults, but on the other hand, these assemblages might also have formed by size-sorting due to currents.

Conclusions

Early late Viséan sedimentary rocks of the Tafilalt (Morocco) contain well-preserved and diverse ammonoid assemblages, which enabled us to emend the diagnoses of the two ammonoid species Goniatites lazarus and Calygirtyoceras darkaouaense. Additionally, we describe the new species Entogonites bucheri sp. nov., which shares the tetragonal juvenile whorls with other representatives of the genus (although the tetragonal shape is only weakly developed in the new species).

The locality SE of Dar Kaoua in the Tafilalt occasionally yields blocks full of small, mostly juvenile specimens. These juveniles commonly exhibit shapes that never occur in adult ammonoids. Also, the morphologic change in the corresponding adults is often profound. We discuss this finding in the light of Pareto Optimality, which was applied to ammonoids by Tendler et al. (2015). In accordance with the results of studies by Jacobs (1992) and Jacobs and Chamberlain (1996), we suggest that the different morphospace occupations of juvenile and adult ammonoids might represent optimisations for compactness (and secondarily hydrodynamics) in the juveniles and for hydrodynamics (and secondarily shell growth) in adults. In future studies, it might be interesting to further compare juvenile and adult morphospace occupations and optima, their degree of overlap and separation as well as changes in overlap through time and phylogeny.