Scarini possess two forms of tight dental organization
The marginal teeth of Scarini are tightly organized into alternating vertical rows (Figure 13). Dentine is highly reduced and do not form a shaft (Figure 15). Functional teeth are located on the margins of the jaws while their successors and dental germs at various developmental stages are located supramarginally (premaxillae) or inframarginally (dentaries) (Figure 13). While scrapping species have relatively homogeneous marginal teeth, marginal teeth of excavating species (Cetoscarus, Bolbometopon and Chlorurus) present a strong antero-posterior gradient of morphology with larger teeth anteriorly. This results in a very tight arrangement mesially while distal parts present more reduced interdental proximity. While all teeth of scrapping species are functional, only one of every two teeth is functional in excavating resulting in a crenate cutting margin. Contrary to Cetoscarus and Bolbometopon, the marginal teeth of Hipposcarus, Chlorurus and Scarus are entirely covered by a layer of bone-like tissue .
Sparisoma present diverse arrangement types
In Sparisomatini, coalesced dentitions are restricted to Leptoscarus and Sparisoma genera (Figure 16A). In these two genera, premaxillae are always coalesced with also a bone-like cover of the marginal teeth but dentaries vary in term of coalescence since some dentaries are not coalesced (Fig. S 5). As Scarini, dental successors are respectively located supra- or inframarginally to the functional dentition in premaxillae and dentaries. However, the number of developing teeth is greatly reduced resulting in apparent oblique rows. Marginal teeth can present various degree of mesio-distal gradient of morphology similarly to Scarini. Leptoscarus possess relatively similar teeth while some Sparisoma haves a strong anterio-posterior gradient of tooth morphology (Figure 16B).
Loss of dentine in tightly-organized dentitions
In non-coalesced dentitions such as Calotomus (Figure 17A), marginal teeth are constituted of a long dentine shaft topped by a large enameloid cap (Figure 12, Figure 17 A). The attachment is highly similar to what is found in another Labridae, Labrus bergylta (Berkovitz & Shellis, 2017), with the teeth being implanted within a bony cript. Teeth are basally and lateraly ankylosed by attachement tissue. This type of attachment has been defined as thecodont in Berkovitz & Shellis (2017). However, in the case of Calotomus, we prefer the term subthecodont (Bertin et al., 2018) as the walls of the bony crypt differ between lingual and vestibular sides with a reduced wall vestibularly. In Calotomus, the inner wall of dentine delimits the pulpal cavity. The external layer of dentine is coated by a thin layer of a mineralized tissue. This coating tissue could be collar enameloid (Sasagawa & Ishiyama, 1988) or acellular cementum (Soule, 1969). In Leptoscarus or non-coalesced dentaries of Sparisoma, marginal teeth possess a smaller dentine shaft than in Calotomus. The dentine is reduced in paving dentitions (Figure 15) and even more in interlocking dentition, where dentine is restricted to a thin layer under the enameloid cap (Figure 17B). The coating tissue is missing in interlocking- and paving-type teeth (Figure 15, Figure 17B).
Sparisoma and Leptoscarus
In Sparisoma or Leptoscarus (Figure 19B, Figure 20AC, Fig. S 11, Fig. S 12), the marginal dentition is constituted of apparent oblique rows of teeth. There is a fundamental difference between this oblique pattern of tooth arrangement and the biological process responsible for tooth replacement.
When we look at the pattern of marginal tooth arrangement in Sparisoma and Leptoscarus (Figure 19B) (but also in other Sparisomatini Figure 19C), we first had the visual feeling that the teeth are arranged in oblique rows and that each oblique row forms a tooth family ranging from the tooth that is developing up to to the one who is in functional occlusion (Figure 19B). This would imply that the teeth gradually migrate along an oblique constraint from their development until the arrival in function. However, none of the teeth examined in all the dental rows shows any trace of displacement whose motor would be the oblique replacement. Such oblique replacement would imply that the teeth of the same row would show traces of resorption distally to their bases. We had never observed any of these phenotypes. An alternative hypothesis has therefore to be tested to explain the dental replacement process with oblique apparent rows (Figure 19C). In this second hypothesis, it is the lag between the offsets of tooth production that produce the oblique rows. In most dentitions of Sparisoma and Leptoscarus, teeth at similar mineralizing stages are usually distant of three or four tooth positions (Figure 21BC). On the contrary, teeth are produced in an alternating way in Scarini (Figure 21A). These differences of tempos of tooth production strengthen the feeling of oblique patterns in Sparisoma and Leptoscarus compared with Scarini.
We also found other clues that disregard oblique rows as tooth families. Some Sparisoma and Leptoscarus specimens present narrow symphyseal marginal teeth that are easily distinguishable from other marginal teeth (Figure 19, Figure 25H). These teeth are organized into interrupted vertical rows, which prove the existence of vertical replacement in Sparisomatini. Similar vertical replacement can be seen on the mesial side of the dentition of the largest specimens of Sparisoma (Fig. S 12CDFGH) with teeth at similar mineralizing stages that are only distant of two tooth positions similarly to Scarini. If oblique rows were tooth families, there should be no replacement within a row. However, we found such occasional replacements (Figure 25A). In addition, a break in the dental plate that is anterior to the formation of an oblique family should prevent the establishment of the most mesial part of the row (see description in Figure 25B-E). However, we found evidences that it is not the case. Some Sparisomatini also present some non-oblique patterns in their dentition (Fig. S 15). Similarly to Scarini (Figure 21), Sparisomatini have also distal oblique rows that are not associated with mineralizing replacement teeth (Fig. S 12L).
Parrotfish ontogeny: increase of the number of replacement teeth associated with more vertical pattern
We did not found any oral teeth in collected oceanic larvae of Scaridae. Our smaller settled juveniles of Scarus psittacus have five oblique rows resembling the ones of an adult Sparisoma (Figure 26A, A’). The most distal row is constituted of small conical teeth, which mesially extends to about a third of the dentition. The number of functional and mineralizing teeth quickly increases with growth while the distal row of conical teeth is more and more restricted to the distal parts of the dentitions (Figure 26A-D).
As in adults (Figure 19A), oblique lines could still be seen within dentitions (Figure 26A’-D’). Replacement teeth make distinct alternate vertical lines when the specimens get larger (Figure 26C’’D’’). The number of these lines is more numerous in the mesial area as more developing teeth are present. In the largest specimens, clear vertical rows can be seen within dentitions at the mesial side (dark red lines, Figure 26D’’). As a result, the appearance of vertical rows (dark red lines) within Scarus psittacus dentitions is not associated with a progressive verticalisation of the oblique lines (blue lines) as the two geometrical compounds coexist similarly to what is found in adults.
Shift of feeding behaviours is recapitulated by ontogeny
Our oceanic parrotfish larvae do not yet possess an oral dentition, which is implemented during settlement (Bellwood, 1988; Chen, 2002). This may indicate a pelagic suction-feeding.
Our Scarus psittacus settled juveniles already possess dentitions with replacement teeth but some studies describe the first-generation teeth (Bellwood, 1988; Chen, 2002; Sire et al., 2002). These teeth develop extraosseously (Sire et al., 2002) and are organized as small marginal adjacent conical teeth that extend from mesial to distal ends of the cutting margin (Bellwood, 1988; Chen, 2002). With growth, these conical teeth are more and more restricted to the distal side of the dentition (Bellwood, 1988; Chen, 2002). These conical teeth can be observed on the distal ends of the dentitions of our juveniles Scarus psittacus. Such teeth can also be found in other fishes with dental plates such as tetraodonts (Fraser et al., 2012). The reduction in the number of conical teeth is associated to a shift of diet from animal-dominated diet to vegetal-dominated diet (Bellwood, 1988; Chen, 2002). In Scarini, the conical teeth are used after settlement to capture benthic crustaceans or foraminifera (Bellwood, 1988; Chen, 2002). The shift to grazing behaviours goes along with an increasing number of non-conical marginal teeth (Bellwood, 1988; Chen, 2002). The Sparisoma larva has a dentition with pointed teeth that could also be used for predation.
The tempo of tooth production increases in Scarini and Sparisoma in association with grazing
Tooth replacement is highly controlled in teleost fishes and often produces predictable patterns (Smith, 2003; Huysseune & Witten, 2006). The replacement pattern derives from the initiation of the first generation of teeth and is maintained through life as a sort of default state (Huysseune & Witten, 2006). The initiation of first-generation teeth can be set up at adjacent or alternate positions (Smith, 2003; Huysseune & Witten, 2006), and is considered to be driven by the propagation of a molecular signal produced by the first formed tooth (Van der heyden & Huysseune, 2000; Gibert et al., 2019). On the contrary, common variations of replacement pattern suggest that replacement is under local control (Huysseune & Witten, 2006). For instance, zebrafish pattern progressively changes from alternate replacement to a replacement of one of every three teeth. These changes may be explained by difference in the length of the replacement cycle (Huysseune & Witten, 2006). Describing tooth rows is not obvious in parrotfish as both vertical and oblique components can be described in their marginal dentition (Bellwood, 1994) (Figure 21). The predominance of vertical and oblique components also depends on the method used by the observer. The tight organization of Hipposcarus and Scarus marginal teeth is interpreted as vertical rows in Bellwood’s study (external observation) and our study. However, the looser organization of Chlorurus and the paving dentition of Cetoscarus/Bolbometopon (Chapter IA) induced a different interpretation of these dentitions by Bellwood (1994) i.e. an oblique mosaic formed by 45° tooth rows (angle measured in respect to the cutting edge).
In teleost fish, each tooth family is characterized by a different epithelial invagination (Reif, 1982). We showed that oblique rows found in Sparisomatini do not constitute tooth families as defined by Reif (1984) but are rather form by controlled pattern of tooth production. This pattern is close to the ones of captorhinid reptiles (Leblanc & Reisz, 2015) and cichlid Eretmodus (Huysseune et al., 1999). A similar pattern may also be found in some Sparidae (see Fig. 1A in Hughes et al., 1994) or Chaetodontidae (see Fig. 3 in Motta, 1984). Similarly to many Sparisomatini dentaries, teeth erupt vestibularly in Eretmodus resulting in a similar linguo-vestibular orientation of tooth groups. The authors proposed that each group was formed by multiple dental families (Huysseune et al., 1999). The hypothesis stating that tooth groups were formed by a single tooth family would require a strong mesial migration of teeth, which is unlikely (Huysseune et al., 1999). In the case of Sparisomatini, this mesial drift of teeth is even more unlikely since they often present coalescence and tight dental organization (Chapter 1A). We also fail to find any oblique bone resorption similarly to the study of (Huysseune et al., 1999), which could suggest such a movement. The vertical replacement pattern of easy-distinguishable marginal teeth, the discontinuity of some oblique rows and the occasional replacement within a row are also strong evidences of the vertical replacement. As in Eretmodus (Huysseune et al., 1999), the development of teeth of a same family in Sparisomatini can be separated by three to four replacement waves and it is not the new-tooth eruption that leads to the shedding of the previous one. The tooth-shedding mechanism is trivial in Sparisomatini with dental plates since both teeth and bone-like tissue are destroyed by feeding. However, it is less obvious for non-coalesced Sparisomatini since the bony ridge seems less impacted by feeding. The mechanism may be related to bone resorption as it was proposed in Eretmodus (Huysseune et al., 1999).
Table of contents :
LIST OF TABLE AND FIGURES
RESUME SUBSTANTIEL EN FRANCAIS
Origin and functions of teeth
Homology and diversity of teleost teeth
Parrotfishes (Scarinae) are specialized Labridae
Parrotfish are unique among coral reef fishes by their feeding behaviour and their ecological importance
Previous studies on parrotfish dentitions
AIM OF THE THESIS
ORAL TOOTH TERMINOLOGY
CHAPTER 1. MAIN EVOLUTIONARY TRENDS OF PARROTFISH DENTITION
Chapter 1A: the same feeding behavior evolved three times independently in parrotfish from different dental arrangements
Chapter 1B: the evolution of oral dentition in parrotfish is underpinned by different tempos of tooth replacement
Chapter 1C: Phenotyping plicidentine and associated mineralized tissues in oral fangs of parrotfish (Scarinae, Labriformes)
CHAPTER 2. ECOMORPHOLOGY OF THE FEEDING APPARATUS
Chapter 2A: Two independent emergences of grazing in parrotfish associated with dental convergence but osteological differences
Chapter 2B: MicroCT-based determination of digestive contents in coral reef fishes
CHAPTER 3. ECOLOGICAL ROLE OF PARROTFISH IN A DIVERSE FISH COMMUNITY
Chapter 3A: Synchrony patterns reveal different degrees of trophic guild vulnerability after disturbances in a coral reef fish community
GENERAL CONCLUSION & PERSPECTIVES
Emergence of grazing in parrotfish
Ontogeny of pharyngeal dentitions
Is morphological diversity associated to diet specialization?
How dental plates are formed?
Ecological study of fish population