Plicidentine in the oral fangs of parrotfish (Scarinae, Labriformes)

Abstract

Parrotfish play important ecological roles in coral reef and seagrass communities across the globe. Their dentition is a fascinating object of study from an anatomical, functional and evolutionary point of view. Several species maintained non‐interlocked dentition and browse on fleshy algae, while others evolved a characteristic beak‐like structure made of a mass of coalesced teeth that they use to scrape or excavate food off hard limestone substrates. While parrotfish use their highly specialized marginal teeth to procure their food, they can also develop a series of large fangs that protrude from the upper jaw, and more rarely from the lower jaw. These peculiar fangs do not participate in the marginal dentition and their function remains unclear. Here we describe the morphology of these fangs and their developmental relationship to the rest of the oral dentition in the marbled parrotfish (Leptoscarus vaigiensis), the star‐eye parrotfish (Calotomus carolinus), and the palenose parrotfish (Scarus psittacus). Through microtomographic and histological analyses, we show that some of these fangs display loosely folded plicidentine along their bases, a feature that has never been reported in parrotfish. Plicidentine is absent from the marginal teeth and is therefore exclusive to the fangs. Parrotfish fangs develop a particular type of simplexodont plicidentine with a pulpal infilling of alveolar bone at later stages of dental ontogeny. The occurrence of plicidentine and evidence of extensive tooth wear, and even breakage, lead us to conclude that the fangs undergo frequent mechanical stress, despite not being used to acquire food. This strong mechanical stress undergone by fangs could be linked either to forced contact with congeners or with the limestone substrate during feeding. Finally, we hypothesize that the presence of plicidentine in parrotfish is not derived from a labrid ancestor, but is probably a recently evolved trait in some parrotfish taxa, which may even have evolved convergently within this subfamily.

Some parrotfish display fangs that exhibit a particular type of simplexodont plicidentine with a pulpal infilling of alveolar bone at later stages of dental ontogeny. The extensive tooth wear and the breakage of certain fangs lead us to conclude that they undergo frequent mechanical stress, although they are not used to acquire food. This strong mechanical stress undergone by fangs could be linked either to forced contact with congeners or with the limestone substrate during feeding. We hypothesize that the presence of plicidentine in parrotfish is not derived from a labrid ancestor, but is probably a recently evolved trait in some parrotfish taxa, which may even have evolved convergently within this subfamily.

1. INTRODUCTION

Parrotfish form a monophyletic group of 10 genera and 100 species living in tropical and subtropical oceans, with maximum species diversity located in the Indo‐Pacific zone (Fricke et al., 2021; Parenti & Randall, 2011; Rocha et al., 2012). Although they have long been considered an independent family, it is now accepted that parrotfish should be classified as a subfamily (Scarinae) nested within the Labridae (wrasses) (Baliga & Law, 2016; Westneat & Alfaro, 2005). Scarine fish live along shallow water, rocky shores, and seagrass beds, almost exclusively in coral reef environments (Froese & Pauly, 2021; Streelman et al., 2002). They are major contributors to bioerosion within coral reef environments, in which they play a pivotal role in the resilience of these highly threatened habitats (Bonaldo et al., 2014; Viviani et al., 2019).

Scarines are currently divided into the two tribes: Sparisomatini and Scarini (Baliga & Law, 2016; Schultz, 1958; Smith et al., 2008; Streelman et al., 2002). The tribe Sparisomatini (Figure 1a) includes the genera Leptoscarus, Calotomus, Nicholsina, Cryptotomus, and Sparisoma, which are generally associated with herbal reef environments as they feed primarily on marine angiosperms or macroalgae (Bellwood, 1994; McClanahan et al., 1999; Nakamura et al., 2006), except for some species of the genus Sparisoma (Bellwood, 1994; Bernardi et al., 2000). In contrast, the tribe Scarini (Figure 1a) includes the genera Cetoscarus, Bolbometopon, Hipposcarus, Scarus, and Chlorurus, which feed on hard substrates, including dead corals and rubbles (Bellwood & Choat, 1990; Bonaldo et al., 2014). Comparative anatomy and feeding behavior data show that scarine fish are microphagous, primarily digesting encrusting cyanobacteria (Clements et al., 2016).

FIGURE 1.

Phylogeny of Scarinae and images of the 3 parrotfish species studied in this article. (a) The phylogeny (adapted from Kazancioğlu et al. (2009) and Baliga & Law (2016)) includes the 10 extant genera of parrotfish grouped within the subfamily Scarinae (red bar) as well as their sister group, the subfamily Cheilinae. Scarine fish are subdivided into two groups: the tribe Sparisomatini (green bar) and the tribe Scarini (blue bar). Parrotfish genera that lack supra‐marginal or distal fangs are marked with an asterisk. (b) Leptoscarus vaigiensis TP, specimen MBIO1853, Standard Length (SL): 245 mm. (c) Calotomus carolinus TP, specimen MBIO1452, SL: 273 mm. (d) Scarus psittacus TP, specimen MBIO890, SL: 200 mm. (b, c, d) Fish photographs were taken by Jeffrey T. Williams, National Museum of Natural History, Smithsonian Institution, and come from the French Polynesia Fish Barcoding Database (http://fishbardb.criobe.pf).

Parrotfish are reputed to be among the most colorful fish in coral reef environments. A majority of parrotfish species have a complex socio‐sexual system, breaking down into three phases, each phase being characterized by a different color pattern (Barlow, 1975; Reinboth, 1968; Thresher, 1984). The first phase comprises sexually immature individuals that are usually dull in color. The second phase or initial phase (IP) includes sexually mature males or females that are very difficult to distinguish because they display a similar color pattern. The third or terminal phase (TP) only includes the famous bright‐colored mature males, which generally dominate breeding activities through a harem‐based social system (Van Rooij et al., 1996).

Parrotfish are also known for their so‐called “beak,” an anatomical feature after which the group takes its common name (Monod et al., 1994). This beak‐like structure forms an extremely compact feeding apparatus at the margins of the oral cavity, with which some parrotfish are famously able to bite into limestone (Bonaldo & Bellwood, 2009; Frydl, 1979). This adaptation for durophagy underlies the importance of parrotfish to bioerosion in coral reef environments (Bellwood et al., 2003; Bruggemann et al., 1996). However, the marginal feeding apparatus of parrotfish is not a keratinous beak like those of birds and turtles, but a combination of bony plates and masses of coalesced teeth.

The oral dentition of parrotfish consists of two subsets of teeth: (1) the marginal teeth that participate in the coalesced bony‐dental structure; and (2) a series of large fangs that protrude from the upper jaw, and more rarely from the lower jaw. These fangs are not present in all species of scarine fish, but among the species that possess them, fangs begin to protrude from the jaws in advanced IP individuals and they are well developed in TP males. This observation led Schultz (1958) and Bellwood (1994) to consider the presence of these fangs as a secondary sexual feature. Considering their unusual anatomical positions (Figures 2, 3, 4), fangs are clearly not involved in feeding, and may instead play another role. Up to now, the developmental and evolutionary origins of these fangs as well as their specific function are poorly understood.

FIGURE 2.

Organization of the premaxillary dentition in an adult male (TP) Leptoscarus vaigiensis, specimen ASM‐2015‐095, SL: 186 mm. (a, b, c) External mesial views of the premaxillaries. (a) 3D volume showing that only the fangs and the functional tip of the marginal dentition are anatomically apparent. (b) Maximum projection rendering providing in situ observation of the number of teeth and their arrangement within the marginal dentition. (c) Tracing of elements of interest along the rendered image in (a). White dotted lines show the parts of the image (a) that have been cropped. Taken together, (a) (b) and (c) show that only the functional part of the marginal dentition appears (in white), while the rest of this dentition is covered with a bony plate (in bright green). Supra‐marginal fangs (in orange and red) are implanted at the dorsal limit of the marginal dentition. The mesial supra‐marginal fangs (in orange) are graphically differentiated from the distal supra‐marginal fangs (in red). The enameloid cap of fangs is drawn in white and the dentine (dark orange and red) is differentiated from the attachment tissues (light orange and red). The black lines drawn on the surfaces of distal fangs correspond to the observable dentine folds. Surfaces colored in dark gray indicate scars from lost or broken teeth. (d) External lateral view of the distal supra‐marginal fangs borne by the left premaxilla. Only the left fang was visible on the frontal view as the right fang is erupting. As the erupting fang is not yet attached to the premaxilla, we were able to virtually extract it from its socket to display it in full on the image (e). (e) Ventral and mesial view of the virtually extracted fang showing the undulations of the outer surface of the dentine which become more pronounced toward the base of the tooth. Headlight symbols indicate the direction of view in the image whose letter is indicated in brackets. Scale bars are 2 mm for (a,b,c,d) and 500 μm for (e).

FIGURE 3.

Organization of the premaxillary dentition in an adult male (TP) Calotomus carolinus, specimen ASM‐2014‐002, SL: 232 mm. (a, b) External mesial views of the premaxillaries. The large mesial marginal teeth (in blue) are distinguishable from the much smaller crowns of the distal marginal teeth, which are covered by a bony plate (in bright green). Two supra‐marginal fangs (in red) are present distally on each side of the upper jaw. Enameloid caps of teeth are drawn in white and the dentine (dark blue and red) is differentiated from the attachment tissues (light blue and red). The black lines drawn on the surface of distal fangs correspond to the observable dentine folds. Dark gray surfaces indicate abrasions or fractures on crowns of the mesial teeth. (c) External lateral view of the left premaxilla, which shows the robust and curved shape of the supra‐marginal fangs. White dotted lines show the parts of the image (a) that have been cropped. (d) Oblique, ventral view of the lower and upper dentition near occlusion. The black dotted circle shows that the lower marginal teeth abut against the dentinal shaft of the first left upper fang, which generates a notch (e) Ventral surface of the second upper left fang which shows numerous folds within the dentinal shaft. Headlight symbols indicate the direction of view in the image whose letter is indicated in brackets. Scale bars are 3 mm for (a, b, c, d) and 1 mm for (d, e).

FIGURE 4.

Organization of the premaxillary dentition in an adult male (TP) Scarus psittacus, specimen AST‐2017‐048, SL: 290 mm. (a, b, c) External mesial views of the premaxillaries. (a) 3D volume showing that only the distal fangs and two rows of marginal teeth are visible. (b) Enlarged view on which the anatomical elements of interest have been schematized. (c) Maximum projection rendering allowing observation of the whole marginal dentition. Taken together, (a) (b) and (c) show that only the functional part of the marginal dentition appears (in white), while the rest of this dentition is covered by a bony plate (in bright green). A transverse row of supra‐marginal fangs (in red) is implanted at the distal limit of the marginal dentition. Enameloid caps of fangs are drawn in white and the dentine (dark red) is differentiated from the attachment tissues (light red). Dark gray surfaces indicate worn or broken fangs. (d, e) 3D volume and associated enlarged diagram of the external lateral view of the left premaxilla. (f, g) 3D volume and associated enlarged diagram of the external lateral view of the left dentary. White dotted lines in (b, e, g) show the parts of the associated images that have been cropped. Color codes for (e) and (g) are the same as for (b). Headlight symbols indicate the direction of view in the image whose letter is indicated in brackets. All scale bars are 2 mm.

In the present article, we examine the finer‐scale anatomy and histology of fangs in sparisomatin and scarin parrotfish. Through detailed CT and histological analyses, we show that some of these fangs bear complex infoldings along the tooth bases called plicidentine, a feature that has never been reported in parrotfish. Plicidentine is the general term for radial folding of mineralized dentine around the pulp cavity at the base of a tooth (Maxwell et al., 2011). Mature dentine is a hard tissue and is therefore incapable of folding. However, during root development, epithelial cells outlining the developing tooth root fold in different directions and guide the dentine‐producing odontoblasts to produce the final folded appearance (Palci et al., 2021). Hence, plicidentine corresponds to a particular shape of dentine and is not to a unique type of dentinal tissue (Peyer, 1968). As the presence of plicidentine has already been extensively documented in many unrelated vertebrate groups (Kearney et al., 2006; Kiprijanoff, 1881; Maxwell et al., 2011; Palci et al., 2021; Ricqlès & Bolt, 1983; Schultze, 1969; Warren & Turner, 2006), it seems likely that plicidentine is homoplastically distributed across vertebrates and serves a functional role in maintaining the integrity of teeth (MacDougall et al., 2014; Palci et al., 2021; Preuschoft et al., 1991). In Actinopterygii, the presence of plicidentine is rather uncommon, but recent work has significantly increased the number of reported cases within this group (Germain et al., 2016; Germain & Meunier, 2017, 2020; Grande, 2010; Meunier et al., 2013; Meunier et al., 2015a). Here we present a detailed study of the plicidentine in parrotfish, which curiously only occurs in teeth that are not associated with the feeding apparatus. We analyzed both the external and internal morphological details of the dentine, its histological features, and its relationships to the surrounding dental and periodontal tissues. Conventional X‐ray microtomography was used to visualize three‐dimensional aspects of plicidentine within the fangs and also to generate virtual slices. In addition to the microtomographic approach, we performed hard tissue histology in areas of greatest interest, which allowed us to document the detailed structure and arrangement of the various mineralized dental tissues and surrounding attachment tissues of parrotfish fangs.

2. MATERIALS AND METHODS

2.1. Specimen sampling, fixation, and housing

We studied 5 marbled parrotfish (Figure 1b), Leptoscarus vaigiensis (Quoy & Gaimard, 1824), 5 star‐eye parrotfish (Figure 1c), Calotomus carolinus (Valenciennes, 1840), and 15 palenose parrotfish (Figure 1d), Scarus psittacus Forsskål, 1775. Specimens were ordered or purchased from local fishermen who collected them by spear fishing around Moorea and Tahiti Islands (Society Archipelago, French Polynesia). After measuring specimens, the heads and digestive contents were dissected and fixed separately in EtoH 70. Specimens were listed in a database containing information on sampling site, standard length, sex and/or phase. The heads were then dehydrated by successive immersion in EtOH baths increasingly concentrated up to 100°. EtoH was then removed with a desiccator to obtain totally mummified heads. This technique preserves the organization of soft and mineralized tissues, their interconnections, and it ultimately yields higher quality 3D reconstructions. At present, specimens are housed in the personal research collections of the team of Laurent Viriot and image stacks for each published specimen are available upon request.

2.2. 3D imaging and virtual histology

Dentition and associated mineralized tissues were scanned from dehydrated heads of parrotfish using conventional X‐ray microcomputed tomography (Phoenix Nanotom S, General Electrics). All specimens were scanned using a 0.1 mm copper filter, 100 kV voltage, and 70 mA current. We obtained image stacks of 3000 successive X‐rays for each sample with an exposure time of 500 milliseconds. After reconstructing 3D volumes by using dedicated Phoenix software, volumes were segmented using VG Studio Max 2.2. Three‐dimensional volumes of jaws were generated with voxel sizes ranging from 10 to 15 μm. In order to characterize dentine structure in more detail, we carried out complementary scanning on smaller parts of the jaws and we generated 3D volumes with voxel sizes ranging from 3 to 6 μm. VG Studio Max software makes it possible to visualize tissue organization in 3D with various renderings or to virtually cut volumes according to any section plan to document internal structures. Here we used the scatter rendering to visualize external features and the maximum projection rendering as well as the virtual slicing mode to visualize internal features and tissue organization.

2.3. Hard tissue histology

A skeletonized premaxilla of Leptoscarus vaigiensis was sectioned for a histological characterization of plicidentine. The sample was first embedded in clear‐setting Castolite polyester resin and placed under vacuum. After the resin had cured, the embedded specimen was mounted to a Buehler Isomet 1000 low‐speed wafering saw and cut along the appropriate sectioning planes. The cut surface of the block was then polished with 600‐ and 1000‐grit silicon carbide powder and mounted to plexiglass slides using Scotchweld SF‐100 cyanoacrylate adhesive. The mounted block was then cut away from the slide, leaving an approximately 150‐μm‐thick wafer mounted to the slide. This wafer was then ground down to optical clarity using a Hillquist grinding machine, as well as 600‐ and 1000‐grit silicon carbide powder. The section was then polished using a fine polishing cloth. The thin section was imaged under plane‐ and cross‐polarized light using a Nikon Eclipse E600 POL polarizing microscope with a Nikon DS‐Fi3 microscope camera and NIS‐Elements‐D imaging software.

3. RESULTS

3.1. General characters of the oral dentition in scarine fish

All scarine fish possess an oral and a pharyngeal dentition, which together perform different but complementary functions. The oral dentition comprises a set of marginal teeth which are placed on the edges of the jaws, at the immediate entrance of the oral cavity (Figures 2b, 3a, and 4c). The lower and upper marginal dentitions do not occlude when the jaws are closed, because the lower dentition bites in front of the upper dentition in Sparisomini, whereas it is the opposite in Scarini (Schultz, 1958). Since the marginal dentition is supplied by a continuous dental replacement system, it gathers not only the functional teeth located on the margins of the jaws, but also their successors as well as dental germs at various developmental stages that are located deep within the dentary and premaxilla (Figures 2b and 4c).

The oral dentition of scarine fish may also include fangs, which are larger teeth that radiate around the vestibular (=labial or external) surfaces of jaw bones (Figures 2, 3, 4). These fangs are mainly present in the upper jaws, they are absent in younger specimens, moderately present in IP individuals, and they reach their maximum development in TP males. As for all dental positions of parrotfish, fangs are replaced periodically. Parrotfish fangs were previously described by Schultz (1958) and Bellwood (1994) as “canines”, a terminology that we did not retain because the canine is a well‐defined tooth of mammals only (Smith & Dodson, 2003). We, therefore, used the word “fang” because it is commonly used without potential homological connotation to designate teeth which are distinguished from other teeth by their protruding, pointed, and possibly curved character.

3.2. Oral dentition of the marbled parrotfish (Leptoscarus vaigiensis)

The marginal dentition of Leptoscarus vaigiensis is composed of numerous obliquely oriented dental rows covered on their vestibular side by a bony plate (Figure 2a‐c). Teeth are, however, not in contact with each other and they do not form an interlocked dental wall, as in Scarus. The marginal teeth have short roots and are composed of an ovoid and vestibulo‐lingually compressed crown with a rounded enameloid cap (Figure 2b). These teeth are all uniform in shape and size, with the exception of the most distal rows, in which teeth are smaller and more pointed.

Two different types of supra‐marginal fangs are present dorsal to the marginal teeth in males only (Figure 2). Mesial supra‐marginal fangs are stocky and conical, whereas distal supra‐marginal fangs are more elongated and dorso‐ventrally compressed (Figure 2c‐e). Compared to the marginal teeth, supra‐marginal fangs have short crowns and long roots. Supra‐marginal fangs are also continuously replaced, but both the number of successors and the replacement rate seem lower than those of the marginal teeth. Consequently, some supra‐marginal teeth may be shed and detectable only by remains of their attachment scars (Figure 2ac). Dental replacement of supra‐marginal and marginal teeth can interfere with each other, causing changes in the geometrical organization of one and/or the other dental group (Figure 2b).

3.3. Oral dentition of the star‐eye parrotfish (Calotomus carolinus)

The marginal dentition of Calotomus carolinus consists of two groups of teeth, the first being located mesially and the second distally (Figure 3a‐c). Mesial marginal teeth are large, with lanceolate and vestibulo‐lingually compressed crowns, lateral cutting edges, and clearly visible roots. Distal marginal teeth are much smaller, with lanceolate to conical crowns and no distinguishable roots (Figure 3c). The distal marginal teeth are roughly organized in mesiodistally aligned rows and their bases are coated by a bony plate (Figure 3bc), similarly to the marginal teeth of Leptoscarus.

Supra‐marginal fangs are present distally to the mesial marginal teeth and dorsally to the distal marginal teeth (Figure 3). These distal supra‐marginal fangs are comparable in size to marginal mesial teeth, but their crowns are conical and pointed, and the whole tooth is curved backwards (Figure 3ce). Since the lower dentition comes in front of the upper dentition when jaws are closing, some fangs display mechanical marks indicating that they act as a stop, which forms a notch at the base of the root of certain functional fangs (Figure 3d). Fangs are often present in advanced IP individuals and are always present in TP individuals.

3.4. Oral dentition of the palenose parrotfish (Scarus psittacus)

The marginal teeth of Scarus psittacus are organized in alternating and interlocked vertical rows, clearly visible by x‐rays of the jaws (Figure 4c). The vast majority of the marginal dentition forms a dental wall entirely coated by bone, together forming a coalesced bony‐dental structure (Figure 4ab). Marginal teeth not participating in the coalesced dental structure form a row of small conical teeth in the distal part of the jaws (Figure 4de). Marginal teeth lack roots and have lanceolate and vestibulo‐lingually compressed crowns with flat tips (Figure 4c). Crown bases are bulged and hollow, including an extremely reduced circumpulpal dentine layer. The hollow bases of marginal teeth allow each successive tooth to fit into the pulp cavity of the preceding tooth, forming interlocking columns of teeth.

Scarus psittacus display large fangs that protrude distally to the coalesced dental structure on the upper and the lower jaw (Figure 4). Among our parrotfish sample from French Polynesia, we detected that some species of Scarus, such as S. altpipinnis, S. niger and S. forsteni, do not develop fangs on the dentary (Viviani & Viriot, personal observations). In Scarus psittacus, fangs stand closer to the marginal zone than in Leptoscarus and Calotomus, but they are slightly staggered with respect to the marginal distal teeth (Figure 4). As fangs do not contact food items when the jaws close, we do not consider them as marginal, but para‐marginal teeth. These para‐marginal fangs of Scarus psittacus have conical crowns and long roots (Figure 4ab). The upper fangs tend to be curved backwards (Figure 4de), while the lower fangs are obliquely oriented either upwards or backwards (Figure 4fg). Fangs are often present in advanced IP individuals and always present in TP individuals.

3.5. Plicidentine in the oral fangs of scarine fish

3.5.1. Surficial features

Based on a detailed examination of jaws reconstructed in 3D, certain external characters hinted at the presence of plicidentine within the fangs, but not within the marginal dentition. In Leptoscarus vaigiensis, the long roots of the distal supra‐marginal fangs show longitudinal ridges, which become more pronounced toward the base of the teeth (Figure 2c‐e). Longitudinal ridges are also visible on the roots of some distal supra‐marginal fangs of Calotomus carolinus (Figure 3de). These ridges are more subtle than those observed in Leptoscarus and they are limited to the mesial sides of the fangs. In contrast, no obvious external axial ridges are detectable along the surface of the para‐marginal fangs in Scarus psittacus.

3.5.2. Virtual histology

Virtual slices were performed through the rendered models of parrotfish jaws (Figures 5 and 6). These slices reveal that vertical grooves are present on the circumpulpal dentinal wall of fangs in Leptoscarus vaigiensis, Calotomus carolinus, and Scarus psittacus (Figure 5ab). These grooves are continuous with the external longitudinal ridges previously observed (Figures 2e and 3e), attesting that folds are present in the entirety of the dentine.

FIGURE 5.

Longitudinal sections of fangs of Calotomus carolinus (a), Leptoscarus vaigiensis (b) and Scarus psittacus (c). These sections show that fangs are shallowly implanted into the alveolar bone sockets. (a) Dentine folds become clearly apparent in the basal third of the inner dentinal wall and the pulp cavity is entirely free of mineralized tissue. (b) Two‐thirds of the inner dentinal wall shows marked dentinal folds and alveolar bone is (modestly) present within the pulp cavity. (c) The dentinal wall is very thick (the dentinal folds will only be visible on perpendicular sections) and a massive bony plug has invaded the pulp cavity. All scale bars are 700 μm.

FIGURE 6.

Internal structure of mineralized tissues in the supra‐marginal and distal fangs. (a, b, c, d, e, f) Virtual horizontal slices through the basal parts of fangs illustrate the diversity of plicidentine patterns as well as the periodontal tissues along the fangs. (a, b, c, d) Sections of four different fangs of Leptoscarus vaigiensis, specimen ASM‐2015‐095. (a) Section of a newly emerged, not yet functional fang showing no contact between the plicidentine (bright green) and the edges of the socket made of jaw bone (light blue). The pulp cavity is free of any mineralized tissue (black circle bordered in white). (b) As the fangs become more mature, their outer dentinal wall first fuses with the jawbone via alveolar bone mineralization (red), while the pulp cavity remains free. At the periphery of the dentine layer is a white border that is consistent with cementum (fuchsia arrow). (c) Once the fang is attached externally, a second phase of alveolar bone mineralization (red) gradually invades the pulp cavity. (c) is an early stage and (d) is a late stage of alveolar bone mineralization of the pulp cavity. (e, f) Fangs of Calotomus carolinus, specimen ASM‐2014‐002, and Scarus psittacus, specimen AST‐2017‐048, respectively. (e) and (f) show a diversity of dentinal wall plication pattern in fangs that are both externally attached but still have a free pulp cavity. All scale bars are 500 μm.

An animated serial section of a fang of Leptoscarus vagiensis (Movie S1) confirms that the dentinal folds are absent from the upper half of the tooth, but gradually become more pronounced toward its base (Figures 6, 7, Movie S1). Drawings of three sections selected from the Movie S1 show that each groove of the circumpulpal dentinal wall corresponds to a ridge in the outer dentinal wall (Figure 7c‐h). The dentine has rather rounded contours at mid‐height of the tooth (Figure 7cf), but these contours become more jagged at the base, each peak corresponding to an external dentinal fold (Figure 7eh). When progressing from the mid‐height of the fang to its base, the overall diameter does not expand much, but the dentine folds become more pronounced (Figure 7fgh). Put together, these data suggest that the number and wavelength of the dentine folds barely change toward the base of the tooth, whereas their amplitudes increase considerably.

FIGURE 7.

Variation in dentine folding pattern in a fang of Leptoscarus vaigiensis, specimen ASM‐2015‐095. (a) The fang as seen from its tip perpendicular to the upper jaw. (b) Virtual longitudinal section of the fang. White dotted lines indicate the three virtual cutting planes studied and refer to the diagrams (c, d, e). In the diagrams (c, d, e, f, g, h), the internal and external walls of the dentine are drawn in red and white, respectively, while the dentine is colored in gray and the pulp cavity in black. (c, d, e) show that the number and wavelength of dentine folds change little toward the base of the tooth, while their amplitudes increase. (f, g, h) are cumulative views of the external contour (f), the internal contour (g), and the two contours of the dentine walls. The observation of (f) and (g) show that the complexity of the dentine plications is greater on the distal side of the tooth. d: distal, v: ventral. Scale bars for (a, b) are 500 μm.

The newly erupted fangs exhibit only a thin layer of mineralized tissue, which appears in light gray on virtual sections (Figure 6a). We interpret this tissue to be primary dentine. Sections of more developmentally mature fangs show that the primary dentine reaches a greater thickness, but never completely fills the pulp cavity at the base of the tooth (Figure 6bc). Still in the more mature fangs, a thin layer of mineralized tissue, which appears more radio‐opaque on the virtual sections (Figure 6b‐d), covers the outer surface of the primary dentin. This latter tissue appears as more compact and/or more densely mineralized because it absorbs more the X‐rays (Movie S1). Given that it coats the outer layer of dentine where the tooth attaches to the jaws, this tissue is most similar to the root cementum reported in amniotes and in other teleosts (Hayes, 1974; LeBlanc & Reisz, 2013; Soule, 1969).

In all three species of parrotfish, fangs are implanted in shallow and slightly asymmetrical sockets (Figure 5), which characterizes a subthecodont implantation (sensu Bertin et al., 2018). The attachment of fangs to the jaws varies with the time from which the tooth is formed (Figure 6). Newly erupted, but not yet functional fangs do not contact the bony edges of the socket (Figure 6a). Later on, fangs fuse to the jaws via progressive mineralization of alveolar bone (“attachment bone” sensu Fink, 1981; “spongy bone” sensu Shellis, 1982; Rosa et al., 2021) between the external dentinal wall and the bony border of the socket (Figure 6b‐f). Once the fang is attached externally, a second phase of alveolar bone mineralization invades the pulp cavity (Figures 5c and 6c), which becomes entirely filled in with bone in the oldest functional teeth (Figure 6d). Intra‐osseous spaces, which are likely to be vascular spaces, are small and numerous within the peri‐dental alveolar bone, whereas they are large and less numerous in pulp (Figure 6d, Movie S1). Based on the centripetal growth of the surrounding attachment tissue, it appears as though alveolar bone forms the principle hard tissue in the periodontium of fangs. The attachment of fangs is therefore achieved through a progressive external ankylosis of the tooth base into shallow alveoli followed by a final mineralization of the pulp.

Fang bases are often surrounded by marginal teeth that are either growing or erupting (Figures 5bc and 6b‐d). Mutual disturbances between marginal teeth and fangs frequently lead to changes in the orientation and position of neighboring marginal teeth as well as to local resorption of mineralized tissues (Figures 5c and 6c).

3.5.3. Hard tissue histology

A histological section through a premaxilla of Leptoscarus vaigiensis yielded similar and complementary results to the microtomographic data. The folds form a smooth, undulating outline to the external surface of the tooth base (Figure 8). As studies of plicidentine in other vertebrates demonstrated (Meunier et al., 2015a; Palci et al., 2021; Preuschoft et al., 1991), the folded shape of basal tooth tissues increases the attachment surface external mineralized tissues. The dentine is avascular and contains straight, relatively unbranched dentine tubules (Figure 8bc), which is typical of orthodentine (Peyer, 1968). The dentine tubules extend through the entire width of the lighter primary, and darker secondary dentine (Figure 8b). In some regions of the tooth sections, the primary dentine contains pronounced growth lines that contour the outer surface of the tooth base (Figure 8bc). These growth lines are spaced approximately 10 μm apart and become subtler toward the center of each tooth.

FIGURE 8.

Thin sections through the bases of two supra‐marginal fangs of a Leptoscarus vaigiensis, specimen ASM‐2015‐095. (a) Cross section of the premaxilla showing two supra‐marginal fangs in transverse section and the marginal dentition in the longitudinal plane. White circles edged in black: free pulp cavity, red dots: alveolar bone, bright green dots: secondary dentine, yellow dots: primary dentine, fuchsia arrows: cementum, light blue dots: jaw bone. Images (b) and (c) are focused on the areas of (a) which are indicated by white dotted rectangles. Scale bars are 500 μm for (a) and 50 μm for (b, c).

The external‐most layers of dentine are coated in a thin (10–12 μm) band of acellular tissue, which is distinct from the underlying dentine (Figure 9). This tissue is consistent with acellular cementum reported in stem and crown amniotes, as well as in some teleosts (Soule, 1969; Hayes, 1974; LeBlanc & Reisz, 2013, LeBlanc et al., 2018). The acellular cementum and the surrounding alveolar bone form a firm ankylosis around each tooth base (Figure 9). The alveolar bone contains numerous vascular spaces and a fibrous matrix that makes the tissue more opaque than the surrounding bone of the jaw under plane‐polarized light (Figure 8a and 9a). A discrete layer of alveolar bone surrounds the bases of each supra‐marginal fang. A scalloped, irregular reversal line separates the outer layers of the alveolar bone from the surrounding hard tissues of the jaw and neighboring teeth, indicating that this tissue is resorbed and re‐deposited with each tooth generation (Figure 9a). The pulp cavity of one of the supra‐marginal fangs is hollow, but the neighboring tooth base is infilled with attachment tissue (Figure 8a). This second tooth base possesses a single large channel through which the surrounding attachment tissue has apparently invaded the pulp chamber and filled the internal surface of the tooth. The contact between the attachment tissue and the dentine is sharp, suggesting that the bone tissue invaded the pulp chamber once the dentine was completely mineralized, which is consistent with the X‐ray microtomographic data. The contact between the two tissues is scalloped in some regions, indicating minor resorption of the internal dentine surface in some areas.

FIGURE 9.

Closeup images of the attachment tissues along the base of a supra‐marginal fang of Leptoscarus vaigiensis in thin section. (a) Plane‐polarized light image of the attachment tissues. Note the fibrous texture of the alveolar bone (red dot) compared to the surrounding jawbone (light blue dot). (b) Cross‐polarized light image of the attachment tissues in (a) highlighting the band of acellular cementum (fuchsia dot) coating the dentine of the supra‐marginal fang. Scale bars are 50 μm.

4. DISCUSSION

4.1. The different types of plicidentine in vertebrates

Plicidentine was discovered long before being named and defined. During the first half of the 19th century, Cuvier (1805), Agassiz (1833‐1843), and Owen (1841), respectively, figured and described folded dentine in the dentitions of Anarhichas and Lepisosteus (actinopterygians), and Mastodonsaurus (early amphibian). Subsequently, Owen (1846) defined six different types of dentin present in fish dentitions, of which three types, namely plici‐dentine, labyrintho‐dentine, and dendro‐dentine, would correspond to three progressive stages of complication of folded dentinal tissues. More than a century later, Schultze (1969, 1970) proposed another terminology also based on an increasing complexity of the folding of dentinal tissues, but also on the presence or absence of bone between the external dentinal folds as well as the presence or absence of mineralized tissue into the pulp cavity. Schultze (1970) grouped all folded dentinal tissues under the name plicidentine and subdivided them into three increasing degrees of complexity: polyplocodont, eusthenodont, and dendrodont, with the intention of using these degrees of complexity to describe trends over sarcopterygian dental evolution. More recently, Meunier et al. (2015a) found that the least degree of plicidentine complexity defined by Schultze (1970) was still too complex to match the pattern observed in the dentition of the actinopterygian Hoplias aimara. Consequently, Meunier et al. (2015a) defined the simplexodont plicidentine, in which the tooth has folded dentinal basal walls with only first‐degree branches, a pulp cavity free for any mineralized tissue, and no bone deposited external to the folds (Meunier et al., 2015a).

4.2. Plicidentine phenotype in scarine fangs

Our results show that Leptoscarus vaigiensis, Calotomus carolinus, and Scarus psittacus develop supra‐marginal or para‐marginal fangs, the basal half of which is characterized by the presence of folded mineralized tissues. Transverse sections through these fangs reveal loosely folded plicidentine (sensu Maxwell et al., 2011). Mantle and circumpulpal dentine are folded, but so is the acellular cementum band coating the tooth roots (Figures 8 and 9). As the mantle dentine is the first tissue to mineralize during odontogenesis, the folding pattern of circumpulpal dentine and coating tissues likely is inherited from the primary structure of the mantle dentine. The plicidentine pattern in scarine fangs most closely resembles the simplexodont type (Meunier et al., 2015a). The dentine folds into sinuous primary branches and the pulp cavity is initially empty in fully erupted fangs. However, unlike in simplexodont plicidentine, the surrounding alveolar bone extends between the gentle folds of the dentine and the pulpal cavity is progressively filled with alveolar bone, which is in contradiction with the definition of Meunier et al. (2015a). The mineralization of alveolar bone into the pulp cavity is secondary and comes late during tooth maturation. We, therefore, assign the loosely folded dentine at the base of parrotfish fangs to a particular type of simplexodont plicidentine with a pulpal infilling of alveolar bone at later stages of dental ontogeny (Figure 8a).

4.3. Putative function of fangs

Given their positions on the jaws, the fangs of parrotfish are neither involved in foraging behavior nor in food selection and processing. We observed parrotfish feeding in situ during successive field seasons in French Polynesia and all our observations show that parrotfish feed by scraping or excavating the substrate exclusively using their marginal dentitions (Viviani & Viriot, personal observations). Consequently, and unlike in many other vertebrate groups where these teeth are present, the presence of fangs and of dentinal folds at their bases cannot be linked to any adaptation to a particular type of diet in parrotfish. Furthermore, the three species of parrotfish that we studied histologically here includes two species that browse on fleshy algae (Leptoscarus vaigiensis and Calotomus carolinus) and a durophagous feeder that forages by scrapping hard substrates (Scarus psittacus). These two types of foraging behavior are so different that it reinforces the idea that the presence of fangs is unrelated to trophic affinities.

Schultz (1958) and Bellwood (1994) suggested that these fangs could constitute a secondary sexual character. From our sampling, we observed that fangs generally begin to develop in advanced IP individuals and are well developed in TP males of several species, which supports Schultz and Bellwood hypothesis. However, the absence of fangs in some genera such as Cetoscarus and Bolbometopon suggests that possessing fangs is not a universal feature of adult parrotfish. Moreover, these fangs are probably not used solely for display as they show signs of tooth abrasion and attrition as well as more severe damage such as sharp crown breaks. As mentioned above, fang attrition is produced by the lower teeth abutting against their ventral surfaces (see Figure 3d). Abrasion and breaks of the dental crowns probably have a different origin. The fangs must regularly contact external elements which gradually wear down the teeth, or even break them when the contact is too violent. We cannot exclude the possibility that the abrasion of the fangs could be caused by fights between TP males (Mumby & Wabnitz, 2002), but it is also possible that the fangs serve as lateral bumpers against corals and rocks so as not to injure the lateral surfaces of the head during feeding. Note that the two hypotheses are not mutually exclusive if they are combined in a certain way. In the first configuration, fangs could have a primary display and fighting function in males, and wear or breakage could be caused by accidental contacts with the substrate during feeding. The reverse configuration that head protection would be the main selective function does not work since females should also have fangs when this is not the case. These various hypotheses remain to be tested by behavioral studies filmed in underwater video. Regardless of their specific function, their relatively deep implantation and complex root shapes compared to the marginal teeth would suggest that fangs undergo some mechanical stresses.

4.4. Potential function of plicidentine in scarine fangs

Plicidentine in parrotfish is unusual in that it is only associated with teeth that are not involved with feeding. The remaining oral teeth of the parrotfish we examined either have no roots, or small conical roots with no plicidentine (Figures 2, 3, 4, 5). Since plicidentine is mainly present in large conical fang‐shaped teeth of vertebrates, many authors have interpreted the presence of plicidentine as a property of large kinetic‐feeding predators (Meunier et al., 2013; Modesto & Reisz, 2008; Scanlon & Lee, 2002). This interpretation was recently questioned regarding the presence of plicidentine in teeth of Latimeria chalumnae (Meunier et al., 2015b). As Latimeria feeds through suction, the physical stresses exerted on teeth are low and therefore the presence of plicidentine cannot be related to an improved biting function in that taxon (Meunier et al., 2015b). This functional interpretation is also refuted in parrotfish since their fangs do not intervene in their feeding behavior.

We observed that the backward curved fangs of Calotomus carolinus show more dentinal folds on their mesial face than on their distal face (Figure 6e). Due to their curved shape and considering the direction of movement of the fish, an external stress can only be exerted on the mesial face of these fangs. When a stress is exerted on the whole tooth, most of the pressure results in the mesial part of the fang basis, where the folds are dominant. From this observation, it therefore appears that plicidentine at the base of parrotfish fangs could be linked to a better resistance of the attachment to an external stress, the source of which is still unknown.

We also noted that dentinal folds are only found in large fangs when the same individual has both large and small fangs. This observation was made in Scarus psittacus, in which dentinal folds are absent in small distal para‐marginal fangs. The presence of plicidentine only in the large fangs in parrotfish could be a consequence of their relatively shallow implantation (Figure 5). Preuschoft et al. (1991) explored the problem of having large fangs in small jaws. The development of large fangs normally requires that the jaws have deep sockets in which the teeth fully develop before they erupt. When it is not possible to have such deep sockets in the jaws, which is the case of marginal teeth around fangs in parrotfish, then large fangs erupt, while the tooth is still immature. Preuschoft et al. (1991) concluded that the adaptive response to implanting large fangs in shallow alveoli is to have plicidentine, which allows the immature tooth to retain optimal mechanical properties. A negative relation between the depth of tooth implantation and the presence of dentine folds already was proposed from studies of plicidentine in teeth of parareptiles (MacDougall et al., 2014), non‐mammalian synapsids (Brink et al., 2014), and even snakes (Palci et al., 2021).

4.5. The evolutionary significance of plicidentine in parrotfish

The presence of plicidentine in actinopterygians has long been understudied (Schultze, 1969; Tomes, 1878; Wyman, 1843). However, a recent series of studies shows that plicidentine is present in the dentition of some Polypteriformes (Germain & Meunier, 2017), extinct Saurichthyiformes (Argyriou et al., 2018), Amiiformes (Germain & Meunier, 2017), Lepisosteiformes (Germain et al., 2016; Grande, 2010), Osteoglossiformes (Germain et al., 2016; Meunier et al., 2013), Characiformes (Germain et al., 2016; Meunier et al., 2013; Monod, 1950; Thomasset, 1930), Spariformes (Germain & Meunier, 2020), Lophiiformes (Germain et al., 2016), and Perciformes (Germain et al., 2016; Meunier & Germain, 2018). Through the present publication, we can add occurrences in Labriformes to this growing list of plicidentine cases in actinopterygians. Although the number of orders concerned may seem large, the presence of plicidentine in Actinopterygii represents less than 20 species out of the circa 33,200 species gathered in this superclass (Betancur‐R et al., 2017; Froese & Pauly, 2021). The number of reported species that have plicidentine is therefore very low so far and their distribution highly scattered within extant actinopterygians. There is no doubt that the presence of plicidentine constitutes a convergent dental trait in actinopterygians, as in other vertebrate groups, and this trait is not suitable for studying evolutionary relationships in actinopterygians as well.

With the exception of Lepisosteus, which have fangs exhibiting polyplocodont plicidentine (Meunier & Brito, 2017), all the other actinopterygians possessing teeth with folded dentine display simplexodont plicidentine. Thus, the dominant type of plicidentine in actinopterygians are teeth with loosely folded dentine at their bases, non‐mineralized pulp cavity, and no bone mineralization external to the folds. In parrotfish however, even if plicidentine is effectively simplexodont, we observed alveolar bone mineralization into the pulp cavity, which makes the reported phenotype slightly different from that of the other groups of actinopterygians. On a larger scale, loosely folded dentine is also the dominant phenotype in many groups of derived amniotes (Brink et al., 2014; MacDougall et al., 2014; Maxwell et al., 2011; Palci et al., 2021). It therefore appears that the more complex types of plicidentine are dominant only in basal sarcopterygians and early amphibians (Schultze, 1970).

The homology of plicidentine within the Scarinae subfamily is an important example of the difficulty of attributing the presence of plicidentine in a single group of fish. Even though all the large fangs we examined show plicidentine at their bases, it is possible that they are not homologous at the scale of the Scarinae because although they resemble each other morphologically, they display slight dissimilarities and they do not always occupy exactly the same positions on jaws of the different species. The supra‐marginal fangs of Leptoscarus vaigiensis and Calotomus carolinus could be homologous because Leptoscarus and Calotomus are sister genera. Under these conditions, it is also possible that possessing plicidentine could be a shared trait of these two genera. However, it is impossible to determine whether the fangs of Leptoscarus and Scarus are homologous because fangs have different shapes and different locations on the jaws and the fangs located on the dentary of some Scarus have no equivalent in Leptoscarus and Calotomus. So plicidentine may well have appeared several times independently, even at the scale of the Scarinae subfamily.

In addition, we have studied a few species belonging to the Cheilinae subfamily, which is the closest subfamily of the Scarinae within the Labridae (according to Baliga & Law, 2016). Cheilines feed on a wide range of hard‐shelled invertebrates (Khalaf Allah, 2013; Randall et al., 1978) and they have a dentition with many marginal fangs, which is very different from the feeding habits and dentition of scarines. We did not find in cheilines any sign of plicidentine at the base of their fangs (Figure S1). This observation suggests that plicidentine appeared within scarines and is not derived from a more distant ancestor, but an exhaustive study of plicidentine within all labrid fish will certainly make it possible to assess the current situation much more precisely.

5. CONCLUSION

We propose that the function of plicidentine in the fangs of parrotfish is not related to their feeding habits, but to a mechanism of strengthening the dental attachment when fangs are still immature by increasing the contact surface between tooth and bone in shallow alveoli. Later, when the fangs are fully mature, the invasion of alveolar bone into the pulp cavity could provide a device to improve the strength of the attachment for teeth that appear to be subjected to many external pressures over their functional period.

Overall, the precise pattern of plicidentine as well as the arrangement of the surrounding attachment tissues therefore seems to be carefully adapted on a case‐by‐case basis not only in the different species of parrotfish, but also for the rest of the sufficiently documented examples in vertebrates. This could explain why it is so difficult to classify the different types of plicidentine because they are intrinsically all different.

AUTHORS’ CONTRIBUTIONS

RG, DL, and LV designed the study. RG, VL, TM, VR, JV, and LV collected and prepared the material. JV and AL acquired the data. JV, AL, and LV analyzed and interpreted the data and wrote the manuscript. All authors critically revised the manuscript and approved the current version of the manuscript.

Supporting information

Figure S1

Movie S1

Viviani, J. , LeBlanc, A. , Rurua, V. , Mou, T. , Liao, V. & Lecchini, D. et al. (2022) Plicidentine in the oral fangs of parrotfish (Scarinae, Labriformes). Journal of Anatomy, 241, 601–615. Available from: 10.1111/joa.13673

DATA AVAILABILITY STATEMENT

I declare that all the new data composing this article are available