Tooth serration morphologies in the genus Machimosaurus (Crocodylomorpha, Thalattosuchia) from the Late Jurassic of Europe

Machimosaurus was a large-bodied durophagous/chelonivorous genus of teleosaurid crocodylomorph that lived in shallow marine and brackish ecosystems during the Late Jurassic. Among teleosaurids, Machimosaurus and its sister taxon ‘Steneosaurus’ obtusidens are characterized by having foreshortened rostra, proportionally enlarged supratemporal fenestrae and blunt teeth with numerous apicobasal ridges and a shorter anastomosed ridged pattern in the apical region. A recent study on ‘S.’ obtusidens dentition found both true denticles and false serrations (enamel ridges which contact the carinae). Here, we comprehensively describe and figure the dentition of Machimosaurus, and find that Machimosaurus buffetauti and Machimosaurus hugii have four types of serration or serration-like structures, including both denticles and false denticles on the carinae. The denticles are irregularly shaped and are not always discrete units, whereas the false denticles caused by the interaction between the superficial enamel ridges and the carinae are restricted to the apical region. Peculiarly, the most ‘denticle-like’ structures are discrete, bulbous units on the apicobasal and apical anastomosed ridges of M. hugii. These ‘pseudo-denticles’ have never, to our knowledge, previously been reported among crocodylomorphs, and their precise function is unclear. They may have increased the surface area of the apical region and/or strengthened the enamel, both of which would have been advantageous for a durophagous taxon feeding on hard objects such as turtles.


Summary
Machimosaurus was a large-bodied durophagous/chelonivorous genus of teleosaurid crocodylomorph that lived in shallow marine and brackish ecosystems during the Late Jurassic. Among teleosaurids, Machimosaurus and its sister taxon 'Steneosaurus' obtusidens are characterized by having foreshortened rostra, proportionally enlarged supratemporal fenestrae and blunt teeth with numerous apicobasal ridges and a shorter anastomosed ridged pattern in the apical region. A recent study on 'S.' obtusidens dentition found both true denticles and false serrations (enamel ridges which contact the carinae). Here, we comprehensively describe and figure the dentition of Machimosaurus, and find that Machimosaurus buffetauti and Machimosaurus hugii have four types of serration or serrationlike structures, including both denticles and false denticles on the carinae. The denticles are irregularly shaped and are not always discrete units, whereas the false denticles caused by the interaction between the superficial enamel ridges and the carinae are restricted to the apical region. Peculiarly, the most 'denticlelike' structures are discrete, bulbous units on the apicobasal and apical anastomosed ridges of M. hugii. These 'pseudodenticles' have never, to our knowledge, previously been reported among crocodylomorphs, and their precise function is unclear.   [2]. The specimen herein described is now registered as NHMUK PV R36793, while the second tooth is still within the Lepage collection. The tooth is single cusped, conical, with little labiolingual compression, and is poorly lingually curved (figure 1). The apex is blunt and rounded, and shows evidence of being worn and polished. No apicobasal facets are evident on either the labial or lingual faces (such as those in the metriorhynchid genus Geosaurus [15,17]). Cingula (a ridge at the base of the crown) and accessory cusps/denticles are absent. The tooth also lacks carinae.
In addition to the blunt-apices, another characteristic feature of Machimosaurus dentition is the enamel ornamentation. Tooth enamel ornamentation varies along the crown; in the basal-mid regions, enamel ornamentation is composed of numerous apicobasally aligned ridges of high relief, which in the apical region shifts to an anastomosed pattern (figure 1). This general enamel ornamentation pattern is also seen in 'S.' obtusidens [1,13] and the metriorhynchid To. carpenteri [15,16] (these taxa also have blunt apices). Some of the apicobasally aligned enamel ridges, especially on the lingual surface, are 'wavy' or undulating. Where these undulations are very pronounced, they can be mistaken for serrations when seen without optical aids.           No apicobasal facets are evident on either the labial or lingual faces, and cingula and accessory cusps are absent. There is no evidence of carinal (mesiodistal) wear (such as in the metriorhynchid genus Dakosaurus [18]).
While the only in situ teeth in the holotype (SMNS 91415) are anterior-mid dentary teeth (figure 2), although there are some isolated crowns preserved in matrix, a referred specimen from Ain, France has in situ upper and lower jaw dentition [9]. This specimen shows heterodonty, with decreasing tooth apicobasal length along the tooth row. Isolated teeth referred to M. buffetauti from the Kimmeridgian of France and Germany (figures 3-5) [12], and cf. M. buffetauti from the Kimmeridgian of England (figure 6) [3] show the same variation in tooth apicobasal length. Interestingly, the anterior teeth of the holotype and the isolated, elongate teeth from France and Germany have carinae (figures 2-5), whereas the presumably posterior, shorter tooth from England lacks carinae ( figure 6). This could suggest that presence of the carinae varies along the tooth row. Unfortunately, the in situ teeth of the Ain skull are still partially encased in matrix, so their carinae are not always visible [9]. The tooth crowns which are carinated have a mesial and a distal carina, and do not have split or supernumerary carinae (a carina that is split into multiple forks or multiple carinae located near each other [19]).
The tooth crowns are lingually curved, although this curvature is much more pronounced in the elongate tooth crowns (figures 2-5) than in the presumably posterior, shorter teeth (figure 6). Fortunately, the curvature of in situ Ain skull teeth is visible [9], and this specimen shows the same trend shown by the isolated tooth crowns: stronger lingual curvature of the elongate anterior teeth, and weaker lingual curvature of the shorter posterior teeth. Interestingly, some of the isolated tooth crowns preserved in matrix of the holotype (SMNS 91415) also show a shorter apicobasal length and are more poorly curved lingually than those in situ anterior-middle dentary teeth. Therefore, contra Martin & Vincent [6], the dentition is not homodont.
As with Machimosaurus sp., tooth enamel ornamentation varies along the crown. In the basal-mid regions, enamel ornamentation is composed of numerous apicobasally aligned ridges of high relief, which in the apical region shifts to an anastomosed pattern (figures 3-6). In M. buffetauti teeth, on the lingual tooth surface, the apicobasally aligned enamel ridges immediately adjacent to the apical anastomosed region are closely packed, whereas on the labial surface, these ridges are more widely spaced (figures [3][4][5]. This is especially marked in the French tooth ( figure 5). In the apical-half of the tooth crown, the apicobasally aligned enamel ridges are 'wavy' or undulating. Where these undulations are pronounced they closely resemble serrations (figures 3g, 4g, 5g,h and 6e).
Macrophotographic study of isolated M. buffetauti teeth shows true denticles on the carinae (figure 3g). These denticles are very similar to those seen on 'S.' obtusidens carinae [13], as they are microscopic, poorly defined and difficult to observe with the naked eye (table 1). They are present on both the mesial and distal carinae, but they do not proceed along the entire carina (heterogeneous), but rather appear in short rows. Note that some of the serrations in the apical region are, in fact, false denticles, created by the short anastomosed ridges contacting the carinae (figure 3g and table 2).      In the apical half of the tooth crown, the apicobasally aligned enamel ridges are 'wavy' or undulating, but these undulations are notably less pronounced than those seen in M. buffetauti. Macrophotographic study of isolated M. hugii teeth shows true denticles on the carinae (figures 8b, 11f , 12f and 16a). These denticles are very similar to those of M. buffetauti and 'S.' obtusidens [13], in that they are microscopic, poorly defined, and are difficult to observe (table 1). They are present on both the mesial and distal carinae, but they do not proceed along the entire carina (heterogeneous), but rather appear in short rows. As with M. buffetauti, some of the serrations in the apical region are in fact false serrations, created by the short anastomosed ridges contacting the carinae (figures 8b and 16a).  One of the most curious features of the isolated M. hugii teeth is seen along the apicobasal and anastomosed ridges. On these ridges are what look like accessory denticles (best seen in figures [8][9][10][11][12][13][14][15]. These structures are more numerous, larger and better defined than the true denticles and false serrations seen on the carinae, such that they can be seen with the naked eye, without optical aids, in good lighting. These 'pseudo-denticles' are restricted to the enamel ridges on the apical half of the tooth crowns.

Co-occurrence of true denticles and false denticles
The co-occurrence of both true denticles and false denticles (superficial enamel contacting the carinal keel) on the carinae of M. buffetauti and M. hugii teeth is notable, but not unique to these species. These two serration morphologies also co-occur in the teleosaurid 'S.' obtusidens and the metriorhynchid To. carpenteri [15,16], and possibly also in the 'pholidosaurid'/elosuchid Elosuchus cherifiensis ([20], MNHN.F MRS 340 and other currently undescribed MNHN Elosuchus specimens). All of these taxa share: robust conical teeth, blunt apices and an enamel ornamentation pattern that in the apical region transitions from being apicobasally aligned to an anastomosed pattern. It is in this anastomosed region that the enamel ridges contact the carinal keel and forms false denticles. It has been hypothesized that the co-occurrence of these morphologies may be linked to functional ecology and diet, specifically a durophagous/generalized macrophagous diet in which both crushing and flesh-slicing are important  [13]. It is an interesting hypothesis, which study on durophagous taxa from other clades (such as mosasaurids and other crocodylomorph clades) could help test. Regardless of the precise reason why true denticles and false denticles occur together, the fact that they are found together in a small group of teleosaurids and metriorhynchids suggests that these animals (Machimosaurus spp., 'S.' obtusidens, Torvoneustes) had a generally similar diet, and may have filled somewhat similar roles in their respective ecosystems.
Apart from the co-occurrence of true and false denticles, most of these characteristics (heterodont dentition, the posterior-most teeth having a robust, conical shape with blunt apices, false denticles and an enamel ornamentation pattern that in the apical region transitions from being apicobasally aligned to an anastomosed pattern) are found in other marine crocodylomorphs. These include, most notably, the dyrosaurid Phosphatosaurus gavialoides [12] (and cf. Phosphatosaurus MNHN.F APH 23) and the tomistomine crocodylian Maroccosuchus zennaroi ( [21], MNHN.F APH 18). As such, this dentition type evolved repeatedly during crocodylomorph evolution.

Pseudo-denticles
Perhaps the single most striking feature of any of the Machimosaurus teeth described here is the 'pseudodenticles' on the apicobasal and apical anastomosed ridges of M. hugii crowns. Their presence helps to readily differentiate the teeth of this taxon from other blunt-toothed teleosaurids (such as 'S.' obtusidens, M. buffetauti and Machimosaurus sp.). These are large, well-defined and immediately notable structures, but their potential function is unclear. The 'pseudo-denticles' increase the surface area of the apical region of the tooth crowns. This could be a way of maximizing friction, and therefore facilitating grip, on wet prey items (such as marine turtles, some of which have embedded M. hugii teeth and/or bite marks consistent with M. hugii [4]). These structures could have also have strengthened the enamel.
It is interesting to note that many other durophagous vertebrates have structures that are broadly similar to the 'pseudo-denticles' of M. hugii, although usually not as large and well defined. These include, for example, various mosasaurids [22], bony fishes such as Paralbula [23], crocodylomorphs such as Brachychampsa [24,25], and some mammals (such as the 'condylarth' Periptychus [26]). All of these taxa possess complex and often anastomosing enamel ridges that undulate, sometimes with bulbous projections like the 'pseudo-denticles'. A broader comparative study or biomechanical study is beyond the scope of this paper, but it is possible that these similar structures served to increase surface area for crushing, strengthened the tooth enamel to better withstand hard-objecting feeding, or both (as has been hypothesized for durophagous mosasaurids [27]). There is some support for the increased friction hypothesis from testing on shoe tread-groove design, which found that tread design influences friction, especially in the presence of liquids [28]. However, relating these treadgroove results to tooth crown 'pseudo-denticles' and the 'interpseudo-denticular sulci' would require biomechanical testing.

Conclusion
Here, we comprehensively describe and figure the dentition of two of the most distinctive species of Late Jurassic marine crocodylomorphs, the teleosaurids M. buffetauti and M. hugii (and described an isolated Machimosaurus sp. tooth). We find that both species had heterodont dentition with a variety of enamel ornamentation patterns, including four distinct types of serrations or serration-like structures. These findings therefore increase the roster of serrated taxa within Teleosauridae, a group that was long thought to lack tooth serrations. Machimosaurus buffetauti and M. hugii share the same suite of morphologies as the closely related species 'S.' obtusidens. All these taxa have true denticles and false denticles co-occurring on the same teeth. As such, there is now a subclade of teleosaurids which have: proportionally enlarged supratemporal fenestrae, foreshortened rostra, reduced tooth counts, teeth with blunt apices and heavily ornamented dentition (apicobasally aligned ridges in the basal-mid regions and shorter anastomosed ridges apically) with serrated carinae. The discovery of serrated dentition in Machimosaurus strengthens the contention that the 'S.' obtusidens-Machimosaurus subclade may have had a diet broader than simple durophagy, with flesh slicing also part of their feeding repertoire.
The most curious discovery is the presence of 'pseudo-denticles' on the enamel ridges of M. hugii teeth. This morphology has never been described for a marine crocodylomorph to the best of our knowledge. These 'pseudo-denticles' are better defined and more numerous than the true denticles along the carinae. The presence of this morphology could be a way to increase the surface area of the apical region of the crowns for maximizing friction (which would be advantageous for maximizing grip), or may have strengthened the tooth enamel. However, these hypotheses require biomechanical testing. Nevertheless, the presence of this peculiar morphology demonstrates that thalattosuchian dentition still yields surprises, despite the long-held assumption of dental conservatism in this group.