The dentition of the Late Jurassic dwarf sauropod Europasaurus holgeri from northern Germany: ontogeny, function, and implications for a rhamphotheca-like structure in Sauropoda

Locality and taphonomy of Europasaurus jaw and dental remains

The Langenberg Quarry is comprised of late Oxfordian to late Kimmeridgian limestone beds that crop out at the northern rim of the Harz Mountains near Goslar, Germany (Lotze, 1968; Pape, 1970; Fischer, 1991). Despite their deposition in a marine environment, some beds contain terrestrial vertebrate remains including lizards, atoposaurid crocodyliforms, pterosaurs, sauropod dinosaurs, several taxa of theropod dinosaurs, and multituberculate mammals (e.g., Fastnacht, 2005; Sander et al., 2006; Wings & Sander, 2012; Richter et al., 2013; Carballido & Sander, 2014; Gerke & Wings, 2014; Marpmann et al., 2015; Lallensack et al., 2015; Wings, 2015; Schwarz, Raddatz & Wings, 2017; Martin, 2018). Europasaurus constitutes more than 99% of the dinosaur material discovered in bed 83 of Pape (1970) of the Langenberg section, and the taxon is only known from this bed (Scheil & Sander, 2017). The abundant and exquisitely three-dimensionally preserved material varies from isolated bones to associated partial skeletons of individuals widely differing in ontogenetic stage. So far, no fully articulated material has been reported (see Carballido & Sander, 2014; Marpmann et al., 2015; Scheil & Sander, 2017; Carballido et al., 2019).

Review of the evolution of the sauropodomorph dentition

Within Sauropodomorpha (Upchurch, 1995; Sander, 1997; Barrett & Upchurch, 2005, 2007), the simplest teeth are found in non-sauropod sauropodomorphs, which are blade-like and flattened labiolingually with an expanded crown (Galton, 1985; Sander, 1997; Galton & Upchurch, 2004), with coarse denticles along the carinae (Galton, 1985; Sander, 1997). Both, the labial side and the lingual side are convex, and the lingual side shows a pronounced ridge from the base of the crown over its entire length.

Non-neosauropod sauropods have more complex teeth of a diversity of crown shapes and often show extensive enamel wrinkling (see below). Within the more derived, i.e., neosauropod, sauropods, two main tooth morphotypes, broad-crowned and narrow-crowned, can be distinguished (Huene, 1926; Janensch, 1929, 1935–1936; Upchurch, 1995, 1998; Sander, 1997; Wilson & Sereno, 1998; Upchurch & Barrett, 2000; Chatterjee & Zheng, 2002; Wilson, 2002; Upchurch, Barrett & Dodson, 2004; Barrett & Upchurch, 2005; Chure et al., 2010; Sander et al., 2011).

Broad-crowned teeth are spoon- or shovel-shaped and are characteristic of basal Eusauropoda and basal Macronaria (Janensch, 1935–1936; Upchurch, 1995, 1998; Sander, 1997; Wilson & Sereno, 1998; Sanz et al., 1999; Upchurch & Barrett, 2000; Chatterjee & Zheng, 2002; Wilson, 2002; Upchurch, Barrett & Dodson, 2004; Barrett & Upchurch, 2005, 2007; Chure et al., 2010; Sander et al., 2011; Wiersma & Sander, 2017). The other morphotype, narrow-crowned teeth, also described as pencil- or peg-shaped, are generally less robust than broad-crowned teeth (Sander, 1997; Upchurch, 1998; Upchurch & Barrett, 2000; Barrett & Upchurch, 2005). They evolved convergently at least twice within the Diplodocoidea and derived Titanosauriformes (Calvo & Salgado, 1995; Upchurch, 1995; 1998; Salgado, Coria & Calvo, 1997; Sander, 1997; Wilson & Sereno, 1998; Sanz et al., 1999; Christiansen, 2000; Upchurch & Barrett, 2000; Curry Rogers & Forster, 2001; Wilson, 2002; Upchurch, Barrett & Dodson, 2004; Barrett & Upchurch, 2005; Chure et al., 2010; Sander et al., 2010, 2011; Wiersma & Sander, 2017).

Alternative, more elaborate classifications than broad-crowned and narrow-crowned have also been used in sauropodomorphs, such as spoon-like teeth, peg-like teeth, cone-chisel, and chisel- like teeth (Calvo, 1994), vulcanodontid teeth (leaf-shaped), spatulate teeth (in Euhelopus, Shunosaurus, Mamenchisauridae), mesiodistally expanded spatulate teeth (in Camarasaurus), cone-chisel like teeth (in Brachiosauridae), slender cone-chisel like teeth (in Titanosauridae), peg-like teeth (in Nemegtosauridae and Diplodocoidea) (Upchurch & Barrett, 2000), leaf-like teeth (the non-sauropod Anchisaurus), spoon-shaped teeth (in Kotasaurus, Barapasaurus), spatulate teeth (in Shunosauridae, Mamenchisauridae, Euhelopodidae, Brachiosauridae, Patagosaurus), teeth with a nearly circular diameter (in Diplodocoidea, Titanosauridae) (Stevens & Parrish, 2005).

Sauropod teeth generally possess a distinctly wrinkled enamel surface, and different patterns of wrinkling have been described with the aim of distinguishing taxa and clades (e.g., Holwerda, Pol & Rauhut, 2015; Mocho et al., 2017; Holwerda et al., 2018; Bindellini & Dal Sasso, 2021). Wrinkled enamel was recognized as a synapomorphy of Eusauropoda (Wilson & Sereno, 1998; Allain & Aquesbi, 2008; Holwerda, Pol & Rauhut, 2015; Carballido et al., 2017; Wiersma & Sander, 2017). This feature, together with the ITRs, may indicate the presence of soft tissue holding the teeth in place after death (Wiersma & Sander, 2017).

Clearly, understanding tooth morphology and jaw morphology, and reconstructing soft part anatomy are of crucial paleobiological importance for understanding sauropods as living animals (Sander & Clauss, 2008; Klein et al., 2011; Sander et al., 2011). Sauropods are unique in the ratio of skull mass to total body mass, and their entire energy uptake must have occurred through those relatively minute skulls. Extreme adaptations thus should not come as a surprise, and major and detailed efforts in understanding their jaw function and feeding adaptations are wholly justified (e.g., Whitlock, 2017).

Material

The majority of the Europasaurus bones and teeth, including a well preserved partial skull (Laven, 2001; Marpmann et al., 2015) was discovered between 1998 and 2002. Because the locality is an active quarry and the strata are steeply inclined (slightly overturned), only limited systematic excavations were possible. Instead, rock stockpiles were screened after every blasting, and fossil-bearing blocks were collected. Currently, more than 1,300 bones and teeth of Europasaurus, belonging to at least 21 individuals of different sizes and ages have been excavated (Scheil & Sander, 2017). The fossils are accessioned to the collections of the Verein zur Förderung der niedersächsischen Paläontologie (FV) which currently is housed at the DfmMh, resulting in the collections acronym DfmMh/FV.

Our study is based on a rich and diverse sample. It consists of a variety of well-preserved isolated tooth-bearing jaw elements of different ontogenetic stages (Marpmann et al., 2015), two ITRs (DfmMh/FV 580.1 and DfmMh/FV 896.7), a partially articulated skull (original SNHM-2207-R and its cast from an earlier preparation stage, NLMH 105.996), and 90 isolated teeth recovered from the matrix of the fossiliferous blocks (see above on collection history). All isolated teeth are well-preserved and lack indications of post-mortem transport. Most teeth were found isolated, in the micritic limestone matrix, only a few were situated adjacent to bones. Two more Europasaurus ITRs have been found but could not be studied: one ITR in the DfmMh/FV collection had been lost in a fire, and another ITR is part of a private collection (Marcus Schipplick, Braunschweig).

The isolated tooth-bearing jaw elements investigated are a total of 13 dentaries (DfmMh/FV 033, DfmMh/FV 034, DfmMh/FV 059, DfmMh/FV 092, DfmMh/FV 093, DfmMh/FV 094, DfmMh/FV 290, DfmMh/FV 291.11-holotype, DfmMh/FV 501, DfmMh/FV 653, DfmMh/FV 654, DfmMh/FV 834.7, and DfmMh/FP 1058.14), five premaxillae (DfmMh/FV 32, DfmMh/FV 61, DfmMh/FV 652.2, DfmMh/FV 982, and DfmMh/FV 291.18-holotype), but only one maxilla (DfmMh/FV 291.17-holotype) (see also Marpmann et al., 2015). Note that the isolated jaw bones consistently lack any functional teeth which apparently were prone to postmortem loss from their respective alveoli, being preserved individually or as the ITRs. The partially articulated skull (SNHM-2207-R, cast NLMH 106.996) preserves the right and the left dentary, a part of the left maxilla, and a large part of the dentition. This skull also documents the seemingly loose implantation of the teeth in the jaws and their rapid post-mortem loss from the alveoli. Thus, we investigated dental wear in isolated teeth as well. We note that it was not possible to assign the material described in this study to the two size classes apparent in the skull (Marpmann et al., 2015) and axial skeleton (Carballido & Sander, 2014) of Europasaurus holgeri.

For comparison, the following material was studied first hand: Giraffatitan specimens from the Tendaguru Formation (MFN t1, now MB.R.2223; S66, now MB.R.2180; S116, now MB.R.2181; and WJ 4170, now MB.R. 2390), Camarasaurus sp. specimen SMA 0002 ‘ET’ (Wiersma & Sander, 2017), holotype specimen of Kaatedocus siberi SMA 0004 (Tschopp, Mateus & Benson, 2015), and Galeamopus sp. specimen SMA 0011 ‘Max’ (Tschopp, Mateus & Benson, 2015) from the Late Jurassic Morrison Formation.

Methods

Morphological examinations were performed macroscopically by sight and microscopically under a stereo microscope. An electronic caliper (measurement error +/− 0.5 mm) was used, and the total preserved length as well as the crown length was measured for all teeth. The length of the crown was measured along the longitudinal axis from the crown tip to the base of the enamel cap. The width of the crown was measured along the widest mesiodistal expansion. These measurements were used to calculate the slenderness index (SI), which represents the ratio between crown height and crown width (Barrett & Upchurch, 2005). We follow the usage of Upchurch (1998), Upchurch & Barrett (2005), and Chure et al. (2010) in that broad-crowned teeth have an SI of ≤4.0, and narrow-crowned teeth have an SI of ≥4.0. We note that other cut-offs have been used (e.g., D’Emic, 2012; Mannion et al., 2013).

The terminologies used in dentistry, in the descriptions of reptile teeth (Edmund, 1969; Smith & Dodson, 2003), and in the descriptions of sauropod dentitions (Wiersma & Sander, 2017) were used in this study.

The wear stages of the teeth were described according to Janensch (1935–1936), Chatterjee & Zheng (2002, 2005), and Wiersma & Sander (2017). Three wear facets were described by these authors and are used in this study (Fig. 1). First, the terminal (or occlusal) wear facet occurs only at the tip of the crown, possesses a round to oval shape, and is steeply inclined. We prefer “terminal” over “occlusal” because of the purely descriptive nature of the former term. The second wear facet is the elongated main wear facet that occurs on one of the carinae of the crown (lower jaw: distal; upper jaw: mesial, Fig. 1). This wear facet is larger than the terminal wear facet. The side wear facet (Fig. 1) is similar to the main wear facet, but is located on the opposite carina of the crown (lower jaw: mesial; upper jaw: distal), and it is usually not as well developed as the main wear facet (Fig. 1).

Based on features such as the degree of wear, root length, tooth exposure, tooth length, tooth size, and the relative position in the alveoli, we assigned developmental stages to the dentition of Europasaurus using the approach of Chatterjee & Zheng (2005). F1 to F5 are stages of wear of functional teeth, and R0 to R3 are stages of development of the replacement teeth (Table 1).

The developmental and wear stages of the teeth were plotted on charts and used in determining Z-spacing (Edmund, 1969; DeMar, 1972; Osborn, 1970; Whitlock & Richman, 2013; Schwarz et al., 2015). In Z-spacing, the number of tooth positions between two teeth in the same stage of development is measured. Since all preserved dentaries contain only replacement teeth, and the functional dentary dentition is only represented by a partial tooth row in skull SNHM-2207-R and an incomplete ITR DfmMh/FV 896.7, the Z-spacing diagrams for the dentary cannot be regarded as error-free and only indicate a general trend.

Tooth formula

Based on the number of alveoli in the jaw bones (Table 2), the tooth formula for Europasaurus is 4 pm + 12 m/13-14 d, as described by Marpmann et al. (2015). All premaxillae clearly show four alveoli, and the only preseved maxilla 12 alveoli. Of the complete dentaries, three (DfmMh/FV 033, DfmMh/FV 034, DfmMh/FV 653) show 13 alveoli. The other complete dentary (DfmMh/FV 291.11), which forms part of the holotype, is slightly larger than the others and shows 14 alveoli. The variability in dentary alveolar counts has been described for other sauropods (Stein et al., 2010), and could represent intraspecific variation. Moore et al. (2018) had suggested that there might be an ontogenetic reduction in the tooth count in sauropods, but there is no evidence for this in Europasaurus despite the nearly three-fold size increase represented by the dentaries (see also Marpmann et al., 2015).

Replacement dentition within the jaw bones

Premaxillary replacement teeth

Four of the five premaxillary bones (DfmMh/FV 032, DfmMh/FV 061, DfmMh/FV 652.3, and DfmMh/FV 982) preserve the entire alveolar section (Fig. 2). The premaxilla DfmMh/FV 982 (Fig. 2D) represents the earliest ontogenetic stage (Table 2). The lingual margin of the alveoli is formed by interdental plates, which are completely fused. The replacement teeth are seen in the windows between the fused interdental plates, and it appears as if the teeth form in crypts separate from the alveoli of the functional teeth.

No functional teeth are contained in the premaxillae. The replacement teeth are triangular and slightly asymmetrical in shape, wherein the mesial carina is C-shaped and curved slightly lingually, whereas the distal carina extends almost straight. The labial side is markedly convex, and the tip is strongly curved toward the lingual side. The lingual side of the teeth is slightly concave, and the teeth have a sharp, blade-like appearance. The enamel shows no peculiarities; it is black and has the typical fine wrinkling resembling wrinkling morphotype III of Holwerda, Pol & Rauhut (2015). The wrinkling appears in the higher developmental stages of replacement teeth (Fig. 2).

Maxillary replacement teeth

Only one almost complete maxilla of Europasaurus is known (DfmMh/FV 291.17), which is part of the holotype (Sander et al., 2006; Marpmann et al., 2015) (Table 2). This maxilla lacks any functional teeth. The alveoli are restricted lingually, and the interdental plates are completely fused. The first three and last six alveoli of the maxilla are clearly seen on the lingual side (Fig. 3), although preservation is not very good. The dimensions of the maxilla and the average distance between the two alveoli indicate that three to four alveoli must have been present in the gap between the two well-preserved parts of the maxilla. We assume a count of three missing alveoli but cannot exclude the possibility that there were four.

Assuming that there were 12 maxillary teeth, replacement teeth are found in almost all preserved alveoli, i.e., in positions 1–3, 7, 8, and 10–12, with alveoli 4–6 not preserved (Table 2). The size of the alveoli and the replacement teeth steadily decrease from the mesial to the distal tooth positions. The labial side of the replacement teeth is convex. Lingually, the mesial teeth have a convex surface, whereas the distal teeth have a slightly concave surface. The enamel of the replacement teeth is smooth, and the enamel caps are very pronounced and round, whereas the carinae are not well developed at this early tooth developmental state.

The eighth alveolus contains the most developed maxillary replacement tooth with black enamel and a fine wrinkling pattern (Fig. 3A, Table 2). It is almost completely erupted and has, in contrast to other mesial teeth, an asymmetrical shape. In lingual view, the crown tip is positioned far distally so that the overall shape of the tooth crown appears more rectangular than triangular. The replacement tooth found in the last alveolus shows an even more pronounced asymmetry. Generally, the asymmetry increases from the mesial to the distal tooth positions, also seen in the dentary tooth row. Maxillary teeth differ from the dentary teeth in that the long axis of the tooth is curved lingually, producing a C-shape.

Dentary replacement teeth

Of the 13 dentaries of different ontogenetic stages, four are complete, and nine are incompletely preserved (Fig. 4, Table 2). The dentaries represent an approximately three-fold size increase from smallest to largest (Fig. 4). No functional teeth are contained in the dentaries. Almost all replacement teeth are in the bell stage without a deeply and firmly anchored root, but bear an enamel cap. The enamel is dark, almost black, and lacks wrinkling. Only the most advanced replacement teeth (R3) that protrude beyond the dorsal margin of the dentary have a slight wrinkling of the enamel, and the root can be identified.

The crown of the replacement teeth is triangular in shape in labiolingual view and slightly asymmetrical. The mesial carina extends further basally than the distal carina and forms a distinct curve, while the distal carina is straight. The visible part of the labial surface is strongly convex, whereas the lingual surface has weak, elongated concavities at the carinae, which are separated by a convex bulge occurring in the middle of the lingual surface of the tooth crown.

In very immature teeth (tooth crown only), which are even more deeply seated in the jaw, this lingual concavity hardly appears, or occurs only on the mesial carina. In further developed teeth, both the mesial and the distal carina are bordered by the longitudinal groove. The tip of the teeth is slightly inclined lingually. The replacement teeth in the distal alveoli differ in shape from the replacement teeth of the more mesial alveoli. Both the mesial and distal carina are quite extensive; the tip is slightly flattened and slightly inclined lingually. The distal teeth show increasing asymmetry (Fig. 4). The carinae and the tips of immature replacement teeth are smooth and rounded compared to those of later developmental stages.

The size of replacement teeth in the dentaries and the diameter of the alveoli decrease steadily from the mesial to the distal tooth positions. A striking feature of the dentaries is that the front replacement teeth (alveoli 1–4) have a strong rotation about their long axis (apicobasal axis) of about 45° from the lateral margin of the jaw, with the mesial carina rotated lingually and the distal carina rotated labially. This long axis rotation decreases posteriorly. The replacement teeth in alveoli 5–8 are only rotated by 15°, whereas the replacement teeth of the rear positions are positioned in parallel to the long axis of the jaw, with the mesial carinae oriented fully mesially and the distal carinae oriented fully distally. The teeth of the dentaries also show a distinctive ‘en echelon’ pattern, i.e., closely spaced teeth in which the distal carina labially overlaps the mesial carina of the succeeding tooth (Wilson, 2002) (Fig. 4).

Almost all replacement teeth in the dentaries show clear and very well formed denticles, mostly only on the mesial carina, but also partly on both carinae. Only well-developed teeth in stage R3, which have already developed wrinkling on the enamel, show no or only very indistinct denticles on the mesial carina. This suggests that denticles are lost as enamel thickness increases during tooth growth.

Very immature replacement teeth located mesially show mostly coarse denticles, five to six in number, whereas older replacement teeth have only three to four poorly defined denticles on the mesial and distal sides. In dentary DfmMh/FV 291.11, the mesial replacement teeth have no denticles. In contrast, both mesial and distal sides of the distal replacement teeth have denticles (Fig. 4A). In dentary DfmMh/FV 654, the replacement teeth in positions 3, 6, and 9 are of approximately the same stage of development and can thus be easily compared. The tooth in the third position shows only very indistinct denticles on the mesial carina, whereas the tooth in the sixth position has five distinct denticles. For the ninth tooth, six denticles are present both on the mesial and distal sides.

A pathology occurs on the fifth replacement tooth in the dentary DfmMh/FV 291.11 (Fig. 4A). This tooth is rotated by 180° in the alveolus, its convex side facing lingually.

Dentition of partially articulated skull SNHM-2207-R

Both the original (SNHM-2207-R) and a cast of an earlier preparation stage (NLMH 105 996) of this partially articulated skull were studied (Fig. 5). In addition to various skull bones (Marpmann et al., 2015), both dentaries and the posterior part of the left maxilla are preserved.

The fragmentary left maxilla has a preserved length of 54.1 mm. All the maxillary teeth are present, yet most are not well preserved, and the crowns are only fragmentary. Although well preserved, the last two teeth are not in their alveoli, but are embedded in sediment nearby. They show a marked crown asymmetry in labiolingual view which is best described as S-shaped. The base of the crown is set at an angle to the root, and the tip of the crown similarly is angled against the more basal part, but in the opposite direction (Fig. 5). Teeth 3 and 6 are replacement teeth of stage R3. The remaining teeth are functional teeth that protrude further from the jaw. As with the teeth in the dentary, the tooth necks are exposed, and the transition from root to crown is about 5 mm above the lateral mandibular margin (Fig. 5).

The left dentary is 124.5 mm long and 40.1 mm high in the region of the third alveolus. Thirteen alveoli can be recognized, but only 12 clearly identifiable teeth are present. The roots of the teeth are circular to oval in cross section. Apical from the crown base, the cross section of the tooth is lens-shaped. The crown expands mesiodistally, whereas the mesial carina always extends further basally than the distal carina. The lingual side is flattened, whereas the labial side is convex. Every tooth has a slightly asymmetrical appearance, which increases from the mesial to the distal tooth positions (Fig. 5). A longitudinal bulge or ridge on the labial side is mesially shifted in the direction of the carina, whereas the top of the crown and the crown base are moved distally.

Independent of stage, the size of the teeth in the left dentary decrease steadily from mesial to the distal. The first four dentary teeth are still preserved in the alveoli, but are strongly shifted posterolingually. The second tooth is very immature, only the crown being visible (stage R3), and it has not reached the occlusal line of the tooth row yet. The other three teeth, located in the alveoli, protrude much further from the jaw. The transition between the crown and root in these three teeth is about 6 mm above the labial mandibular margin, and the neck of the teeth is exposed.

The first tooth in the left dentary shows a slightly inclined terminal wear facet on the crown tip towards the labial side. Toward the distal carina, the main wear facet is advanced, though partially covered in sediment (stage F3). The situation is similar in the crown of the third tooth. It also shows an almost horizontally directed terminal wear facet, a labial patch of polished enamel, and a significant main wear facet. On this tooth, an incipient side wear facet on the mesial carina is also observed. This tooth shows stage F3–4. Teeth 5–10 are displaced from the alveoli but retain their proper arrangement in the sediment, thus forming a partial ITR. The last two teeth, number 12 and 13, are still firmly anchored in the alveoli, although they were reoriented obliquely and anteriorly by taphonomic processes. Their tooth crowns are damaged.

The pattern of tooth replacement is from the distal to the mesial tooth positions, starting at the 10th alveolus. The 5th, 8th, and 11th alveoli contain replacement teeth in stage R2. The tips of these teeth are already far advanced into the alveolus, but have not yet reached the margin of the jaw. The teeth show stages F3–4, and the roots of the corresponding functional teeth 5 and 8 show advanced resorption, being on the verge of replacement by new teeth. The functional tooth from alveolus 11 is missing.

The right dentary is not completely freed from the sediment, and bone fragments cover it. It is 122.6 mm long, and six associated teeth are located nearby. Five of them, which are completely detached from the alveoli, consist only of fragments. The single tooth remaining in the alveoli is a replacement tooth of stage R3.

Isolated upper tooth row DfmMh/FV 580.1

ITR DfmMh/FV 580.1 contains 13 functional teeth and covers a total area of about 100 × 60 mm (Table 3). Comparison with the partially articulated skull indicates that the ITR must belong to an upper jaw (Fig. 6). The longitudinal axes of all the teeth in this ITR are more or less parallel to each other, so it can be assumed that the teeth are preserved in the original position. The teeth show the same morphology as the teeth of the partially articulated skull, and the location of the symphysis in the ITR runs between the eighth and ninth preserved tooth, counted from the left along the ITR in labial view (Fig. 6). This means that the tooth row includes the first left maxillary tooth, the four left and right premaxillary teeth, and the first four right maxillary teeth.

In DfmMh/FV 580.1, each individual tooth is rotated by 20–30° from the tooth row axis, creating an ‘en echelon’ pattern. The mesial carina of the teeth points lingually, and the distal carina points labially, indicating that the mesial side of the following tooth labially overlaps the distal side of the preceding tooth (Fig. 6).

The teeth are exposed from their convex labial side, while the lingual side and sometimes the carinae of the teeth are still covered by sediment. The longitudinal axis of the teeth is strongly curved labiolingually, approximately forming a C shape, because the apex of the crown and the tip of the root point lingually. The roots of the teeth in this ITR are resorbed to varying degrees, and the tips and carinae of the teeth are considerably abraded.

Some of the teeth of this tooth row show a large depression in the enamel, a few millimeters in size, on the labial side in the upper part of the crown. These depressions are pronounced, especially in the first right premaxillary tooth, first and second right maxillary teeth, first and second left premaxillary teeth, and first left maxillary tooth.

In addition, the roots of the teeth show a peculiarity; in labial view, the longitudinal axis of the root runs vertically from the base of the crown to the root tip, but in the last section of the root, the longitudinal axis forms a slight S-curve with the mesially pointed root tip (Fig. 6). This S-shaped configuration of the root tip in labiolingual view is particularly pronounced in the fourth, right premaxillary tooth and the first and third, right maxillary teeth. The crown length and crown width decrease steadily from the mesial to the distal tooth positions, and the S-shaped asymmetry in labiolingual view increases.

Isolated lower tooth row DfmMh/FV 896.7

The ITR DfmMh/FV 896.7 consists of a total of nine associated dentary teeth (Table 3) and two additional, smaller teeth (no number) on a block of sediment. However, the degree of articulation is not as high as in the previously described ITR. The nine teeth are preserved with eight other bones, including several cervical ribs and a cervical vertebra (Fig. 7A). To reveal their morphology in full, the first left dentary tooth (DfmMh/FV 896.7.2, Figs. 1, 7C) and the first right dentary tooth (DfmMh/FV 896.7.3, Fig. 7B) were separated from the block during preparation, but they are considered to belong to this ITR.

This ITR also shows the ‘en echelon’ pattern of the teeth as in ITR DfmMh/FV 580.1 (Figs. 6, 7). As observed in the partially articulated skull SNHM-2207-R and ITR DfmMh/FV 580.1, ITR DfmMh/FV 896.7.1 shows a significant decrease in crown length and width from the mesial to the distal tooth positions, whereas the asymmetry of the teeth increases steadily in distal direction (Fig. 7).

The visible side of the tooth row is strongly curved and thus represents the labial side, whereas the lingual side is hidden in the sediment. The long axis of the teeth is straight, with barely any curvature of the tip to the lingual side of the tooth. These teeth also have an asymmetrical tooth crown, where the mesial carina is extended further than the distal carina, indicating that the symphysis must have been located between the two isolated teeth DfmMh/FV 896.7.2 and DfmMh/FV 896.7.3.

The first to sixth right dentary tooth are most certainly part of the ITR, but the affinity of the two smaller teeth (located about 25 mm away from the rest) is unclear (Fig. 6). The gap between the sixth tooth and the first of the two smaller teeth and the size difference seems to be too large for the two smaller teeth to be consecutive teeth of this ITR.

Isolated teeth

In this section, general individual morphological features and characteristics of isolated teeth are described (Table 4). The isolated teeth (Figs. 1, 8–13) reveal many features, in particular of the lingual side of functional teeth, that cannot be observed fully in the previously described material from skulls and ITRs and hence are described here. A particularly notable morphological feature in 44 teeth is evidence of wear, consisting of enamel polish and tooth wear.

Of the 90 isolated teeth which were studied, 75 are well-preserved. Six teeth are still in matrix blocks and therefore cannot be viewed from all angles. The less well-preserved teeth were damaged during or after discovery, resulting in, e.g., partial or entire lack of the crown.

Denticles

As noted for the in situ teeth of the dentigerous bones, another special feature of replacement teeth are the denticles on the mesial and distal carinae of the crown (Fig. 9, Table 4). Denticles on carinae are a plesiomorphic feature of non-sauropod sauropodomorphs (Barrett & Upchurch, 2007), and are prominent and rough in immature teeth. The further a tooth develops, the more indistinct its denticles become as a result of tooth wear. Among the isolated teeth, those classified as replacement or immature teeth (stage R1 to F1) have visible denticles, and a total of 36 of the 90 isolated teeth show denticles on their mesial carinae.

The isolated juvenile functional tooth DfmMh/FV 860 (Fig. 9C) shows only two denticles. Very indistinct denticles occur on the mesial carina of the teeth DfmMh/FV 422 and DfmMh/FV 731, as well as the teeth DfmMh/FV 460 and DfmMh/FV 707.3.2 (Fig. 9B). Denticles are lacking on the distal carina of the teeth of juvenile individuals, and are completely lacking in large and well developed teeth of adult Europasaurus.

Patterns of tooth wear

Forty-four of the isolated teeth show relatively strong facets on the crown in the form of worn and/or polished enamel (Table 4). The wear facets can be formed and advance differently.

In a very early stage of wear, no facets are visible. The smooth polish spreads from the crown tip onto the labial and lingual sides of the teeth (stage F1). The longer a tooth stays in use, the clearer the grooves and wear facets become. Normally, the terminal wear facet forms first (stage F2). This beginning wear facet has either a round or, more often, a drop-shaped appearance (Fig. 10). The round side of the drop is located on the distal side in the upper jaw and the mesial side on the lower jaw and represents the transition to the side wear facet. The pointed side of the drop thus occurs on the mesial side of the upper jaw and distal side of the lower jaw and transitions to the main wear facet.

In teeth DfmMh/FV 432, DfmMh/FV 440, DfmMh/FV 452, DfmMh/FV 516, DfmMh/FV 537, DfmMh/FV 788, DfmMh/FV 851, and DfmMh/FV 1034.1, the terminal wear facet is rotated from 10 to 50° toward the mesiolingual side (Fig. 10A). Lower jaw teeth DfmMh/FV 439, DfmMh/FV 445, DfmMh/FV 447, DfmMh/FV 456, and DfmMh/FV 896.7.3 show a weakly developed terminal wear facet, which is labiodistally inclined (Fig. 10B).

Two exceptions are DfmMh/FV 461 (Fig. 10C), where the terminal wear facet cuts obliquely onto the distal carina, and DfmMh/FV 436, which has no terminal wear facet, but a very narrow, steeply inclined facet at the distal carina and a slightly larger wear facet on the convex labial side (Fig. 10C). The unusual form of the wear facets of DfmMh/FV 461 and DfmMh/FV 436 could be due to a misalignment of the teeth in the jaw during their lifetime.

As wear progresses further, the main wear facet is usually formed very quickly (stage F3). It is a narrow, elongated facet, which extends from the crown tip to the corresponding carina (Fig. 11). The facet is always slightly concave, caused by the differential wear of the softer dentin over the harder enamel. The main wear facet is usually associated with and develops from the terminal wear facet. Exceptions are again tooth DfmMh/FV 436, which shows no terminal wear facet (Fig. 10C) and DfmMh/FV 851, which has a side wear facet instead of a main wear facet (Fig. 11C), and DfmMh/FV 439 and DfmMh/FV 896.7.3, where the main wear facet is not connected to the terminal wear facet.

As wear progresses, eventually a side wear facet forms (stage F4), but the main wear facet still dominates in size. The side wear facet proceeds deep along the carina, but it is not as wide as the main wear facet.

In rare cases, the side wear facet may eventually surpass the main wear facet in size and spread, as seen in teeth DfmMh/FV 424, DfmMh/FV 452 (Fig. 11C), and DfmMh/FV 455 (Fig. 11A). In the upper jaw teeth, DfmMh/FV 428, DfmMh/FV 454 (Fig. 11B), DfmMh/FV 472 (Fig. 11B), and DfmMh/FV 851 (Fig. 11C), only a small facet on the distal carina has formed below the terminal wear facet.

In the last stage of use, all wear facets fuse and form a joint large wear facet (stage F5), which extends over the entire width of the crown, as in the teeth DfmMh/FV 424 (Fig. 11A) and DfmMh/FV 455 (Fig. 11A). The terminal wear facet disappears as it is integrated directly into the combined main and side wear facets. The enamel band borders the entire wear facet and is raised above the dentin. Interestingly, tooth DfmMh/FV 730 shows a level wear facet with a very irregular outline. It is plausible that the crown had been broken off when the tooth was still in use, and the fracture surface was smoothed over by wear.

On some maxillary teeth with advanced wear, the main and side wear facets are not long, narrow facets along the carina, but rather incorpotated into the teardrop-shaped terminal wear facet. The pointed end of the teardrop shape is thus shifted more and more to the side of the mesial carina (Figs. 11, 12)

Tooth replacement

Like all sauropods, Europasaurus shows a well-organized tooth replacement (Figs. 2, 4–7, 14). The Z-spacing ranges between 2.5 and 3 and therefore lies within the usual sauropod values (Fig. 14). DeMar (1972) determined a Z-spacing of 1.56–2.8 for most reptiles, which may change during ontogeny. The lower the value, the higher the replacement rate (DeMar, 1972). An increase in Z-spacing is observed from the mesial to the distal tooth positions in Europasaurus, which is explained by the fact that the mesial dentition is more stressed and replacement occurs more frequently, thus exceeding the replacement rate in distal teeth. This is also the case for most reptiles (Edmund, 1960; Fastnacht, 2008).

The average Zahnreihe length in the dentary and ITRs of Europasaurus is 2 to 3 teeth (Fig. 14). Here, however, only part of the dentition is considered in each case (for the dentary, only the replacement dentition; for the ITRs, only the functional dentition). The slope of the Zahnreihe varies in the dentary from 1/2 to 2, and in the ITRs from 1 to 3. In general, however, the Zahnreihe of the functional and replacement teeth are uniform and parallel to each other.

Ontogenetic changes in isolated teeth and dentitions

Due to the many well-preserved mandibles with replacement teeth of different developmental stages, ontogenetic changes in the teeth of Europasaurus can be examined in detail. Juvenile teeth are distinguished from those of adult animals by their overall smaller size. However, tooth size also decreases distally, the smallest distal teeth of adult animals are still quite robust and broadly built and show pronounced wear. Juvenile teeth, on the other hand, are much more delicate and sleek with a very narrow crown, which is only slightly wider than the root. The enamel of the crown in juvenile teeth (Fig. 9) is very thin, barely wrinkled and is similar to that of the replacement teeth.

Assigning teeth to jaw quadrants and position in jaw

Based on observations on the tooth-bearing bones and isolated tooth rows of Europasaurus and on published descriptions of sauropod dentitions (Janensch, 1935–1936; Chatterjee & Zheng, 2005; Wiersma & Sander, 2017), the strong lingual curvature of the upper jaw teeth distinguished them from those of the lower jaw. Taking the middle of the crown as point of reference, the apical part of the crown and the tip of the root are curved lingually in the upper jaw teeth. In mesiodistal view, the tooth long axis of the upper jaw teeth is C-shaped (Fig. 13A). The dentary teeth, on the other hand, show less lingual curvature (Fig. 13B). Also, in the upper jaw teeth, the terminal wear facet is wider and more drop-shaped than in the teeth of the lower jaw. The terminal wear facet of the upper jaw is inclined toward the lingual side, whereas the terminal wear facet of the lower jaw is inclined to the labial side. These features in combination allow a distinction between upper and lower teeth.

In labial or lingual view, the tooth crown shows a variably pronounced asymmetry along the mesial carina, resulting in an S-shape (Fig. 8), as also described for Bellusaurus (Moore et al., 2018). The apex of the tooth is shifted towards distal, as is the crown base. The middle part of the crown is shifted toward mesial. The bulge on the labial side of the crown appears slightly curved with the convex side directed mesially and is always located on the mesial side (Fig. 8). This allows for a distinction between left and right teeth.

To differentiate between mesial and distal teeth, the distally increasing asymmetry can be used (Chatterjee & Zheng, 2005; Kosch, 2014; Kosch et al., 2014; this study).

Early stage and location of tooth formation

In the dentigerous bones of Europasaurus (Figs. 2–4), we observed in lingual and occlusal view that the replacement teeth initially appear to form in a crypt spatially separate from the alveolus of the functional tooth. The functional teeth are generally lost, and the alveoli are empty or contain older replacement teeth. Initial formation of the replacement teeth in a crypt happens deep within the alveolar ramus.

Comparison with other sauropod dentitions

The teeth and dentition of Europasaurus have the same attributes as in other basal Macronaria (Janensch, 1935–1936; White, 1958; Madsen, McIntosh & Berman, 1995; Chatterjee & Zheng, 2005; Chure et al., 2010). There thus are no morphological changes that could be directly linked to evolutionary dwarfing. Comparison with the dentition of two closely related taxa, Camarasaurus and Giraffatitan, reveals clear similarities to Europasaurus. The dental formula of Europasaurus (Marpmann et al., 2015) is similar to that of Giraffatitan with pm4 + m11–13/d12–14 (Janensch, 1935–1936; Christiansen, 2000). Camarasaurus hast a dental formula of pm4 + m8–10/d11–14 (Janensch, 1935–1936; Madsen, McIntosh & Berman, 1995; McIntosh et al., 1996; Christiansen, 2000; Chatterjee & Zheng, 2005; Wiersma & Sander, 2017) and thus the same number of teeth in the dentary and premaxilla as Europasaurus and Giraffatitan, but the number of maxillary teeth is lower. The Z-spacing of all three sauropods have values between 2.0 and 3.0 (Chatterjee & Zheng, 2005; Kosch, 2014; Kosch et al., 2014; this study).

Morphologically, the teeth of Europasaurus are sleeker in appearance than those of Camarasaurus, but are slightly more robust and more closely set than those of Giraffatitan. The SI ranges from 1.0 to 2.0 in Camarasaurus (Wiersma & Sander, 2017) but from 2.5 to 3.0 in Giraffatitan (Chure et al., 2010) and Europasaurus (this study), where Europasaurus scores mainly in the lower range. Europasaurus thus falls in the “broad-crowned” category of teeth (Upchurch, 1998; Upchurch & Barrett, 2005; Chure et al., 2010). The teeth of Camarasaurus are labiolingually flattened and mesiodistally widened (Wiersma & Sander, 2017), more so than in Europasaurus. In Camarasaurus, the rounded ridge on the lingual side is very pronounced and the grooves, paralleling the carinae, are present on both the labial and the lingual side. The teeth are only slightly bent lingually and have an increasing asymmetry towards the distal end of the tooth row.

The teeth of Giraffatitan exhibit a marked asymmetry, which increases posteriorly (Janensch, 1935–1936; V. Régent, 2011, personal observation). They are strongly curved lingually. The lingual side of the tooth crown is concave, and the labial side is convex, resulting in a D-shaped tooth cross section. The labial grooves are often only pronounced along the distal carina. Lingually, a groove is evident along the mesial carina (Janensch, 1935–1936; V. Régent, 2011, personal observation). Camarasaurus is not reported to have denticles on either the replacement teeth or the functional teeth (Upchurch & Barrett, 2000; Chatterjee & Zheng, 2005; Wiersma & Sander, 2017), whereas three teeth of Giraffatitan bear indistinct denticles (Janensch, 1935–1936; Upchurch & Barrett, 2000). Europasaurus shows distinct and coarse denticles on immature tooth crowns. These denticles are worn off in the functional teeth. The ‘en echelon’ pattern of the teeth observed in Europasaurus is a synapomorphy of Eusauropods and has been described in the other sauropods detailed above as well as in others (Wilson & Sereno, 1998; Chatterjee & Zheng, 2005; Sereno & Wilson, 2005; Carballido et al., 2017).

The intermediate morphology of the Europasaurus tooth morphology between Giraffatitan and Camarasaurus thus falls in line with its intermediate skeletal morphology. The intermediate morphology has made it difficult to determine the phylogenetic position of Europasaurus, either as a basal macronarian more derived than Camarasaurus or as a member of Brachiosauridae (Carballido & Sander, 2014; Carballido et al., 2019). Dwarfing may have resulted in stem-ward slippage of the taxon (Carballido et al., 2019). Further research in this matter is warranted, for example the interpretation of the conspicuous denticles in Europasaurus replacement teeth in a phylogenetic context (J. Whitlock, 2023, personal communication).

Tooth wear and food intake

There are two different types of wear in sauropod teeth, first described by Calvo (1994), that also occur in the upper jaw teeth of Europasaurus. Type A wear results in a flat, smooth wear facet (e.g., Figs. 1, 9A, 9B and 10A), which does not, or only very late in wear, encroaches on the carinae (Fig. 15). Here, contrary to most wear facets, the enamel always forms a level surface with the dentin, i.e., the dentin is not eroded out relative to the enamel. In type B of Calvo (1994), however, wear reaches the mesial and distal carinae early, and the facet is V-shaped (Fig. 15). The enamel always is raised above the softer dentin. Type A appears along the entire tooth row but the facet becomes increasingly asymmetrical in the smaller distal teeth, whereas type B primarily occurs in the mesial teeth of Europasaurus.

These two types are less pronounced in the dentary teeth which generally show less wear than the maxillary teeth. Often, only the tip of the crown is affected by wear. Mesial teeth usually have a horizontal terminal wear facet, but both main and side wear facets occur on the carinae. The more asymmetrical the teeth are, the more asymmetrical are the terminal wear facets, being occasionally strongly shifted to the distal carina, for example in teeth DfmMh/FV 445 and DfmMh/FV 461.

It is a special feature of Europasaurus that both types of Calvo’s wear facets are developed (Fig. 15; Table 4). Type A, expressed in Europasaurus as tear-drop shaped terminal wear facets, is subordinately observed in Giraffatitan (Janensch, 1935–1936) and is caused by tooth-to-tooth contact (Greaves, 1973), but also caused by biting on hard and tough food (Janensch, 1935–1936; Upchurch & Barrett, 2000; Grippo, Simring & Schreiner, 2004). This facet is exclusively limited to the top of the crown. The teeth located more posteriorly have more asymmetric facets. This suggests that the teeth of Europasaurus did not interlock, unlike those of Camarasaurus (Calvo, 1994; Upchurch & Barrett, 2000; Chatterjee & Zheng, 2005; Wiersma & Sander, 2017), but rather that the lingual side of an upper jaw tooth met the labial side of its antagonist in the lower jaw (Figs. 1, 10).

Type B, the V-shaped wear facet, occurs mainly in the mesial teeth of Europasaurus. Type B, subtype (Fig. 15), wear facets are the most common in Giraffatitan as well (V. Régent, personal observation; Kosch et al., 2014). These wear facets probably were not caused by tooth-to-tooth contact, but primarily by stripping and snipping off vegetation (Christiansen, 2000). The basal extent of the wear facets along the carinae indicates that they could not have possibly been produced by antagonists, unless there was an interlocking organization of the upper and lower dentition. Arguing against such wear caused by interlocking of teeth is the dense spacing of the teeth in the ITRs, however. On the other hand, in these facets, the enamel band always stands proud of the exposed dentin, which is typical of wear caused by foodstuffs (Grippo, Simring & Schreiner, 2004).

It is plausible that Europasaurus consumed different types of foodstuffs. Its relatively robust and broad-crowned teeth were adapted to snap off small twigs and consume tough vegetation, as in other broad-crowned sauropods (Sander et al., 2010; Gee, 2011; Hummel & Clauss, 2011). The mesial and distal wear facets suggest that Europasaurus also used its dentition like a rake, as also was hypothesized for other sauropods (Christiansen, 2000; Upchurch & Barrett, 2000; Barrett & Upchurch, 2005; Hummel & Clauss, 2011). Europasaurus probably did not perform branch stripping in the narrow sense because it does not show the adaptations that have been hypothesized for this feeding mode, which has been drawn into question even for diplodocid sauropods (Whitlock, 2017).

Microwear studies on the teeth of Camarasaurus showed that a ‘chewing-like’, transverse movement of the jaw took place upon closure (Calvo, 1994; Christiansen, 2000). This lateromedial movement of the jaw would also be an explanation for the drop-shaped wear facet of type A observed in Europasaurus and may indicate that the food was at least partially processed orally, as suggested by some authors for Camarasaurus (Calvo, 1994; Christiansen, 2000; Upchurch & Barrett, 2000).

Isolated tooth rows as evidence for anchoring soft tissue structure

Isolated yet still articulated tooth rows are a peculiar but widespread feature among Sauropoda, not only of Europasaurus, and the anatomical structure that led to their preservation may be an autapomorphy of the group. ITRs have been found in Shunosaurus (Chatterjee & Zheng, 2002), Abydosaurus (Chure et al., 2010), Giraffatitan (Janensch, 1935–1936; Kosch et al., 2014), diplodocoids (Britt, Scheetz & Dangerfield, 2008; Whitlock, Wilson & Lamanna, 2010; Peterson et al., 2022), and titanosaurs (Phuwiangosaurus, V. Suteethorn, 2015, personal communication). Notably, ITRs have not been reported for Camarasaurus to our knowledge, despite the abundance of this taxon in the Morrison Formation. The interpretation of this peculiar kind of ITR preservation is potentially relevant to anatomy and feeding behavior of sauropods. The partially articulated Europasaurus skull SNHM-2207 provides evidence that ITRs of sauropods cannot in all cases be attributed to jaw bone loss due to decay, with only the teeth remaining as fossils (McIntosh, 1981), or to consumption of dentigerous bones by insects (Britt, Scheetz & Dangerfield, 2008). In fact, the specific position and preservation of the isolated upper and lower tooth rows in SNHM-2207 provides strong evidence that the preferential destruction of jaw bones is not a plausible explanation for ITRs at all.

A novel hypothesis as to the nature of the ITRs was recently advanced by Peterson et al. (2022) based on a specimen of Apatosaurus from the Morrison Formation of Wyoming, USA. However, the ITR in question is only represented as a cast in the study of Peterson et al. (2022). Based on this single ITR specimen, these authors hypothesize that sauropods replaced their teeth as entire rows at once. They did not consider other specimens of ITRs, and the evidence from the cast is different from what we report here and what has been described in the other ITRs before (Chatterjee & Zheng, 2002; Chure et al., 2010; Janensch, 1935–1936; Britt, Scheetz & Dangerfield, 2008; Whitlock, Wilson & Lamanna, 2010). Except for Chure et al. (2010), these studies are not cited by Peterson et al. (2022). The peculiar and confusing feature in the ITR cast in Fig. 14 of Peterson et al. (2022) is that the unworn teeth protrude well beyond the worn teeth, making it difficult to understand how tooth wear was incurred in this ITR, and raises concerns that the ITR was incorrectly restored during preparation.

Schäfer (1972) provides taphonomic observations on extant porpoises that may shed light on the issue. Mummification of porpoise cadavers causes intensive dehydration and shrinkage of the gingival connective tissue and periosteum. Shrinkage pulls the teeth from the alveoli but preserves the dentition in its original configuration in the dessicated gum tissue (Schäfer, 1972). If this observation is applied to Europasaurus and other Eusauropoda, it would suggest the presence of some kind of connective tissue that kept the teeth together even after death.