The holotype (i.e., IRSNB R 0001) of the sea turtle Eochelone brabantica consists of a partial skeleton that includes numerous disarticulated elements of the carapace (Fig. 1). Erosive alterations occur in several costal plates, two of which exhibit associated bone remodelling (Figs 1c–f, 3, 4). Of these two, the most prominent mark, both in size and bone remodelling, is located in the left first costal plate of the individual (Fig. 1c, d). This anomaly extends through the external cortex into the cancellous bone. The erosion is characterised to be an amorphous penetrative lesion of about 4.5 cm in length and 2.5 cm in width. It presents smooth margins and evidence of bone remodelling on the inner surface. An asymmetric ridge-shaped bone growth of variable thickness is identified within the base of the erosion (Fig. 1c). This ridge exhibits a smooth contour, with a polished surface in appearance, and is nearly as long as the erosive lesion. In ventral view, the specimen presents an asymmetric bone growth with well-defined rounded margins (Fig. 1d). The dimensions of this structure are 3 cm long, 2.5 cm wide, and it protrudes 0.5 cm past the inner bone surface. The surface of this osseous growth is polished in appearance. The second mark associated with bone remodelling corresponds to a circular erosion located on the anterior margin of the left second costal plate (Fig. 1e, f). The hole, with well-defined lateral margins, is about 1.7 cm in diameter and shows a depth of about 0.3 cm. The erosion presents a hemispherical profile, and a relatively flattened base (Fig. 1e). It penetrates perpendicular to the shell surface and through the external cortex and inner cancellous layers reaching the internal cortex layer of the plate. The bone growth associated with the hole is 1.8 cm long, 1 cm wide, and it extends 0.4 cm over the plate surface. It presents relatively well-defined rounded margins, and an apparently polished surface (Fig. 1f). Cross-sectional images obtained through CT scanning reveal the internal morphology of both anomalies and the associated bone remodelling (Figs 1c–f, 3, 4). CT scans of the second costal plate anomaly display an irregular morphology, with a rounded base of the erosion. Additionally, a hyperdense bone region is visible in certain slices, located around the area of the lesion. This corresponds to newly formed bone, indicative of the healing process (e.g., Fig. 3, b1). Multiplanar and three-dimensional imaging of the circular erosion in the third costal plate further corroborate the macroscopic observations (Fig. 4a, b). These images reveal a subcircular erosion, with a relatively flattened base. In this case, the variations in bone density are less pronounced than those observed in the first costal plate. Nevertheless, a hypodense bone structure is apparent in certain slices (e.g., Fig. 4, a4). The similarities in bone density across slices may suggest a more advanced stage of the healing process, as newly formed tissue typically restores its physical and mechanical properties over time.

Figure 3. 

Cross-sectional slices of the left first costal plate obtained through CT scanning of the carapace of the holotype of Eochelone brabantica (specimen IRSNB R 0001) from the Lutetian (middle Eocene) of the Saint-Remy-Geest locality (Walloon Brabant, Belgium). a–c. 3D models of the ventral (a, b) and lateral (c) views digitized by photogrammetry. a1–a8. Perpendicular view of the hole along the XY plane; b1–b8. Perpendicular view of the hole along the YZ plane; c1–c4. Parallel view of the hole along the XZ plane. Scale bars: 2 cm. The arrow indicates the hypodense bone structure.

Figure 4. 

Cross-sectional slices of the left second costal plate obtained through CT scanning of the holotype of Eochelone brabantica (specimen IRSNB R 0001) from the Lutetian (middle Eocene) of the Saint-Remy-Geest locality (Walloon Brabant, Belgium). a–c. 3D models of the ventral (a, b) and lateral (c) views digitized by photogrammetry. a1–a8. Perpendicular view of the hole along the XY plane. b1–b7. Perpendicular view of the hole along the YZ plane. c1–c4. Parallel view of the hole along the XZ plane. Scale bars: 4 cm (a, b); 2 cm (c). The arrow indicates the hypodense bone structure.

The specimen of Eochelone brabantica presents other shell anomalies on its carapace, without evidence of bone remodelling (Fig. 1b, g–j). Such deviations are characterised by being relatively penetrative (i.e., some of them reaching the inner cortex layer of the bone), and subcircular in morphology, with well-defined margins. These shell erosions range from 0.2 cm to 0.5 cm in diameter. All of them are perpendicular to the plate surface, being distributed without a specific pattern over the plates.

The carapace of the sea turtle Puppigerus camperi (IRSNB R 0004) displays various types of alterations, including abnormal erosion of the shell and superficial marks (Fig. 2). Two regions in the posterior area show abnormal absence of some carapacial areas (Fig. 2c). The first erosive lesion corresponds a healed fracture and extends from the left ninth peripheral plate to the centre of the carapace, at least reaching the suprapygal (Fig. 2d). It shows a smooth and irregular contour. This pathology results in the partial loss of the left ninth peripheral and the left last costal plate, with the width of the erosion decreasing towards the centre of the carapace. The second abnormal erosion affects the right eleventh peripheral and the pygal, leading to partial loss of these plates, also displaying a smooth and irregular contour (Fig. 2f). Pits and linear grooves are scattered across the carapace, without a discernible pattern (Fig. 2b). None of them fully penetrate the external cortical bone layer. Consequently, no bone remodelling is associated with these superficial anomalies. The pits measure approximately 0.2 cm in diameter, while the linear grooves are about 2.5 cm long and 0.5 cm wide.

Circular to subcircular pits and holes resembling those observed in the carapace of the holotype of Eochelone brabantica (Fig. 1) commonly serve as indicators of predation or parasitism in modern sea turtles (Frick et al. 2002; Alfaro et al. 2006; Domènech et al. 2015). Notably, the anomalies observed on the carapace of the holotype specimen (IRSNB R 0001) that can be unequivocally attributed to premortem shell marks are illustrated in Fig. 1c–f. As remarked by Zonneveld et al. (2022), this typology of shell erosions is frequently observed on fossilised turtles, being commonly interpreted as resulting from vertebrate bite marks, parasitism, as well as post-mortem scavenging invertebrates (Erickson 1984; Jiménez-Fuentes 2003; Milàn et al. 2010; Noto et al. 2012: Botfalvai et al. 2014; Myrvold et al. 2018; Gônet et al. 2019; Jagt et al. 2020; De La Garza et al. 2021; Drumheller et al. 2023). In the case of the carapace of IRSNB R 0001, the anomalies identified have been ruled out as resulting from predation by a vertebrate organism. This determination is based on several criteria: the lesions do not follow a specific pattern indicative of predator teeth structure; they lack associated drag-out structures, being well-defined without broken edges (e.g., Fig. 1h, i); and no other typologies of shell lesions, such as punctures, scratches, or mechanical deformations, are associated with those observed in the specimen.

Extant sea turtles can harbour a wide variety of epibionts that can compromise the integrity of their shells, leading to the development of various types of shell lesions (Pfaller et al. 2008). Among the epibionts commonly found on turtle shells and capable of causing lesions are cnidarians (e.g., hydrozoans; Pfaller et al. 2008), mollusks (e.g., boring bivalves; Frazier et al. 1985), arthropods (e.g., barnacles; Caine 1986), platyhelminths (e.g., saltwater leeches; Göpper et al. 2018), and algae (e.g., Rhodophyceae; Abel Sentíes et al. 1999). While some of these epibionts act as commensals and cause minimal damage (i.e., hydrozoans and algae), some may function as parasites (i.e., mollusks, leeches or barnacles), inflicting significant lesions, as those observed in IRSNB R 0001. In this context saltwater leeches and mollusks exhibit inconsistencies that make them unlikely contributors to the anomalies observed in Eochelone brabantica. Saltwater leeches have been identified as contributors to certain infections found on the body and carapaces of sea turtles, resulting in erosive lesions on the skin and shell where they attach (Davies and Chapman 1974; Göpper et al. 2018). However, although the morphology of the lesions caused by saltwater leeches may coincide with those observed on the carapace of Eochelone brabantica (e.g., subcircular erosions with well-defined margins), their size does not match with those produced by these parasites. Saltwater leeches that affect sea turtles (i.e., those belonging to the genus Ozobranchus) range in size from 5 to 30 mm (Apathy 1890; Cordero 1929; Ghosh et al. 1963). While it is theoretically possible that fossil leeches capable of causing such lesions could differ in clade or size from extant Ozobranchus species, there is currently no evidence supporting the existence of larger leeches exhibiting similar parasitic behaviour on sea turtles. The lesions on the shell of IRSNB R 0001, with the smallest measuring 80 mm, far exceed the size range associated with any known leeches, whether fossil or extant. Therefore, based on the best available data, saltwater leeches are unlikely to be responsible for the anomalies observed in Eochelone brabantica. Similarly, molluscan organisms such as boring bivalves (pholadids) have also been reported in sea turtles as contributors to shell pitting (e.g., Sato and Jenkins 2020). However, although the size may align with the anomalies observed in Eochelone brabantica, the morphology of the shell erosions typically associated with boring bivalves does not match with that described in IRSNB R 0001. Specifically, boring bivalves are known to produce clavate or tear-shaped lesions (Bromley 2004; Jenkins et al. 2017; Sato and Jenkins 2020), which differ significantly from the circular to subcircular marks discussed here.

Coronuloidea barnacles (i.e., barnacles that are primarily parasites) are epibiotic organisms that live cemented to inert substrates such as rocks or the external surfaces of various marine vertebrates, such as cetaceans or turtles (Kearn 2005). Currently, these cirripedes are classified into three families (Coronulidae, Chelonibiidae, and Platylepadidae, Ross and Frick 2011; Hayashi et al. 2013), with Chelonibiidae and Platylepadidae being mostly known as turtle barnacles (Ross 1963a, b; Ross and Newman 1967; Caine 1986). The association between barnacles and their chelonian hosts has persisted for more than 30 million years (Collareta et al. 2023). Thus, reports of their activity tracks in the form of attachment scars on several marine turtles in the fossil record are prevalent in the literature (e.g., Hayashi et al. 2013; Janssen et al. 2013; Collareta et al. 2019; Collareta et al. 2022a, b). They are often found partly embedded in the turtle carapace and plastron and can cause minimal stress (e.g., increased surface drag) or more serious damage to the shell, leading to bacterial or fungal infections. These parasites commonly penetrate the bone, thus occupying an intermediate position (i.e., mesoparasites), between ectoparasites and endoparasites (Zonneveld et al. 2022). Depending on the shell shape, the ontogenetic stage, and the anchoring apparatus of the barnacles creating the traces, the resulting scars on the turtle shell may vary in shape from shallow circular and hemispherical pits to deep cylindrical holes (i.e., penetrating through the outer dense bone layer into the more cancellous bone beneath, Seigel 1983) (Zonneveld et al. 2015, 2022; Collareta et al. 2022a, b). In this sense, the shell marks produced by barnacles coincide both in size and morphology with those observed in the carapace of Eochelone brabantica (Fig. 1). Specifically, the morphology of these marks includes smooth internal surfaces within the pits, consistent with bone erosion caused by the attachment structures of barnacles, and well-defined but irregular margins that reflect the invasive anchoring apparatus of the cirripedes. The smaller pits (e.g., Fig. 1g) probably represent non-adult barnacles, whereas the deep perforations (e.g., Fig. 1e) suggest that the cirripedes had enough time to develop and acquire significant size. Similarly, shell lesions with amorphous morphology (Fig. 1c) have been interpreted as a probable association of two or more barnacles causing the development of significant scar tissue or some type of infection produced by these parasites. Additionally, distribution of the multifocal shell marks in the individual of Eochelone brabantica are also compatible with the presence of barnacles, as these are commonly located on the shell without a specific pattern, sometimes grouped in a region of the shell or distributed across all areas. Thus, based on the size, internal texture, and marginal shape of the lesions, as well as their distribution, arthropods, particularly barnacles (crustaceans of the subclass Cirripedia), are the most likely to induce lesions similar to those observed in the carapace of IRSNB R 0001.

The IRSNB R 0004 specimen of Puppigerus camperi exhibits various types of erosive anomalies on its carapace (Fig. 2): pits, grooves, and an abnormal absence of a shell region (i.e., including a healed fracture and the erosion of the posterior-most region of the carapace). Notably, the abnormal absence of two shell regions is the only type of injury that can be unequivocally identified as pre-mortem damage and it is the focus of the present discussion (Fig. 2d, e). Determining a specific cause for the fracture is complex, as numerous factors during the turtle lifetime could have led to this type of shell lesion, making it difficult to attribute it to a concrete cause. In this context, some type of traumatism has been considered as the more probable cause for the healed fracture of IRSNB R 0004 (Fig. 2c, d). Likewise, anomalies similar to that observed in the carapace of Puppigerus camperi are commonly associated with traumatic injury in extant sea turtles. When the origin of such anomalies is unknown, they are typically attributed, by default, to boat strikes or predator attacks (i.e., shark attack) (e.g., Orós et al. 2005; Phelan et al. 2006). In this case, it is considered improbable that a predator attack caused the injury, since, as mentioned above, it would be expected to see associated marks of predation (e.g., mechanical deformation or tooth marks), which are not present in this specimen. Consequently, given the lack of specific indicators elucidating the etiology of this trauma, it is deemed inappropriate to provide a definitive diagnosis for the fracture. As a healing process, the shell can regenerate and remodel when injured, as observed in the case of the healed fracture of IRSNB R 0004. In the literature on extant turtles, it is often noted that open shell fractures in aquatic turtles carry a high risk of infection (Mader 2006). Consequently, it is uncertain whether the fracture observed in the fossil was solely caused by the initial physical trauma, whether the injury was more extensive and we are observing the result of subsequent healing, or if minor injuries subsequently became infected, leading to more extensive areas of necrosis. Regardless, the fracture did not involve any particularly vulnerable parts of the body, as the spinal column does not extend this far caudally in the carapace. Therefore, it did not result in the immediate death of the specimen.

Similar to the healed fracture, the abnormal erosion affecting the right eleventh peripheral and pygal plate (Fig. 2e, f) may be attributed to various external factors. For instance, this erosion could be compatible with a healed fracture caused by some type of trauma or could correspond to a set of healed plates eroded by an epibiont (e.g., a mollusk, Hayward and Ryland 1990) rasping on the shell margins when the turtle was not fully developed, as occurs in some extant sea turtles. In any case, this pathology appears not to have impacted the primary functions of the body. It cannot be confirmed whether this abnormal erosion is associated with the healed fracture discussed above or represents separate anomalies. The available evidence does not allow for a definitive conclusion on whether the anomalies in the posterior region of the carapace have a common origin. However, the most parsimonious explanation suggests that both erosive lesions are likely caused by the same factor, given their similar characteristics and location.

Several superficial pits have been observed scattered across the carapace (Fig. 2b). These marks are consistent with two potential etiologies: small invasive barnacles (see section 4.2 for more details), this area being the most likely region in extant turtles to accumulate such organisms; or by some vertebrate or invertebrate activity during the postmortem process. Regarding the linear grooves (Fig. 2b), they are consistent with marks typically associated with organic processes during the postmortem interval, such as invertebrate activity, solution marks made by plants or microorganisms (Fernández-Jalvo and Andrews 2016). Specifically, these grooves are characterized by their U-shaped profile, smooth contours, lack of ornamentation, and gently convex or slightly flattened cross-sections, which align with the morphology expected from such processes (Fernández-Jalvo and Andrews 2016 and references therein). However, it cannot be definitively confirmed that these grooves did not originate while the organism was alive, thus their precise etiology remains uncertain.