We did not find traumatic alteration in the studied material of the following taxa: Stromaphorus, Hoplophractus, Urotherium, Eosclerocalyptus, Pseudoplophorus and Neosclerocalyptus. This may be explained assuming that these animals did not have the fight behavior, or, if they had, the impacts could not provoke serious damages in the opponent. Secondly, the glyptodonts of these genera were medium sized. According to the hypothesis of the sexual selection, an increase of the body size and accessories like horns or spines are typically results of male-male fighting behavior for mates. Blanco et al. (2009) estimate that the masses of the tail of a Pseudoplophorus and Neosclerocalyptus were, respectively, about 15x and 6x less heavy than Panochthus, and although they bear caudal tube, their lateral and terminal figures are smoot, without the marked tubercles and/or striated surfaces, as in Panochthus and Hoplophorus, and probably did not support developed horny spines.
Although these groups were fully armored with compact carapace and caudal tube, there are few samples or, some cases, they were not well preserved. For Doedicurus, we also did not find new cases of alterations, however, we discuss the specimen studied by Lyddeker (1894) which presents an alteration in the dorsal region of the carapace.
Ante mortem alterations and traumas in carapaces
Alterations observed in fossils may have originated either when the animal was alive (ante mortem) or through taphonomic processes (post mortem). We have ruled out the possibility of post-mortem alterations due to the presence of bone response in all analyzed carapaces. Furthermore, we emphasize that the unequal or localized erosions of the ornamentation provide additional evidence of a non-taphonomic process. For instance, erosion caused by transport, a bioestrinomic process, may produce marks quite similar to those produced by ante mortem events. However, transport would result in some level of disarticulation of the carapace, generating small fragments and the abrasion on the osteoderms would be uniform, likely affecting all sides and both external and internal surfaces (Lima and Porpino 2018).
Circular marks on bones can also be produced by taphonomic processes such as necrophagy by insect larvae, which has been previously described in fossil bones (Dominato et al. 2009; Paes-Neto et al. 2016), or tooth marks made by scavengers. However, necrophagic agents prefer nutritious parts like long bones for their bone marrow or soft tissues instead of osteoderms. Additionally, the holes produced by arthropods are usually too small, ranging in the scale of millimeters, unlike the perforations observed in carapaces here, which vary in centimeters.
Predation behavior could leave marks on the exoskeleton, such as holes, perforations, and depressions, but we dismiss this possibility. The Quaternary glyptodonts for which we described such marks here (e.g., Glyptodon, Panochthus, Doedicurus) were massive animals, characterized by an imposing tail and weighing between 750 to 1,4000 kg (Fariña, et al 1998; Barbosa et al. 2023). Their bony armor likely represented an insurmountable barrier even for large contemporaneous predators, such as saber-toothed cats (Smilodon populator), wolves (Canis nehringi), or bears (Arctotherium angustidens) (see Alexander 2001). In MLP-PV 16–36, the two perforations we observed may suggest a bite from a saber-tooth-car, but the distance between the orifices is approximately 18 cm. In contrast, the distance between the canines of such felid is roughly 10 cm. Some authors suggest that young or small glyptodonts could be prayed by felids (Prevosti and Vizcaíno 2006), but others reject this idea (Bocherens et al. 2016; Lima et al. 2022). In any case, our study focuses on adult glyptodonts, so we will not discuss these hypotheses.
Human predation is another possibility since humans coexisted and interacted with the megafauna. Carlini et al. (2022) recorded six cases of predation on glyptodont by humans, noting that in all cases, they were struck on the skull, specifically on unprotected points uncovered by the cephalic shield. In contrast, direct blows to the carapace were likely avoided due to their ineffectiveness.
The signals observed in the first analyzed fragment belonging to the carapace MLP-PV 98-XII-1-1 indicate a deformation associated with a diastatic fracture (Fig. 1d). Fractures are complete or incomplete bone ruptures (Piermattei et al. 2006). However, in the case of MLP-PV 98-XII-1-1, there are no broken osteoderms; they are disconnected due to the massive impact that separated the syndesmoses and caused the deformation. The perforating aspect and the circular depression suggest that this trauma was caused by a spiky structure striking with high energy in a localized area of the carapace, reminiscent of the armored caudal tube with horny spines, like those in Panochthus.
Although the impact was external in the above fragment, its internal surface suffered more damage. If we assume a blow from a caudal tube is a possible cause of these lesions, we can explain this difference through histology, material strength, and energy dissipation on osteoderms/carapaces. The external surface of an osteoderm consists of a superficial zone composed of compact bone, while the middle zone is formed by trabecular bone, which has a lower mineral density (Hill 2006; Pereira et al. 2014; Araújo and Porpino 2018). The caudal tube transferred its kinetic energy to the carapace, and the direct impact on the solid superficial zone destroyed ornamentation and the syndesmoses. In contrast, the middle zone dissipated energy over a larger area in the deep zone, also causing syndesmoses disruption, resulting in more extensive damage to the internal surface. The disruption of the syndesmoses was more pronounced in H. euphrctus (Fig. 7; Table 1), while the two alterations in Panochthus (MLP-PV 98-XII-1-1) are more subtle. The size of the osteoderms may explain this difference. Panochthus osteoderms are larger in area and thickness than those of Hoplophorus, consequently, there are more connective fibers making the syndesmoses firmer and more resistant to impacts. However, we cannot conclusively determine whether this denser bone structure is an adaptation to endure impacts or if it was formed during the animal's life, potentially indicating a physiological or pathological condition.
The injuries in Glyptodon possibly share the same traumatic cause as those in Panochthus. Although individuals of Glyptodon did not possess a caudal tube, the tail tip bore a massive bunch of osteoderms forming a mini club or short tube (Zurita et al. 2010). Some authors suggest they could have used their tail to strike; however, the tail design is less efficient than the caudal tubes of Doedicurus and Panochthus, resulting in weaker blows (Fariña 1995). We noted many small and medium lesions in the carapaces of Glyptodon, while Panochthus and Hoplophorus presented fewer but broader and more severe lesions. Perhaps individuals of the Glyptodon genus increased the number of blows to compensate for the weakness of the impacts. Except for the fracture in MCN-PV 1920, this could explain the scarcity of fractures or depressions in carapaces of Glyptodon, despite the high quantity of lesions observed here in a single individual, because the carapace retained thickness and resistance to endure weak strikes.
The carapaces MACN-Pv 200 and MACN-Pv 13776 exhibited more extensive ornamentation loss than Panochthus and Hoplophorus, likely indicating another evidence of tail blows. Despite its weakness, the Glyptodon tail surface that impacts the carapace is larger than that of caudal tubes with spines, which concentrated the energy of the blow in reduced points. In MCN-PV 1920, the ornamentation of healed osteoderms with bone callus was preserved, suggesting that the impact on the carapace may have struck a nearby region. Like MLP-PV 98-XII-1-1, unaltered osteoderms next to the bone callus retained their ornamentation.
We observed that trauma recovery differs between the internal and external regions of the carapace. Slight to moderate bone response areas were observed on the external surface of the carapaces in all genera in which we identified lesions. In MCN-PV 1920 (Glyptodon) and mainly in MLP-PV 98-XII-1-1 (Panochthus), there are bone calluses on the internal surface of the affected areas of the carapaces, indicative of more extensive healing processes (Waldron 2009). A bone callus is a normal healing reaction that reconnects bones separated by trauma (Waldron 2009). Bone repair typically begins with the deposition of material to form new collagen fibers and cartilaginous tissue, forming a soft callus (Mescher 2021). Full recovery generally takes approximately 16 weeks (Rothschild and Martin 2006). Radiographs confirmed osteogenesis, as evidenced by the higher bone density, a characteristic of calcified bone callus signaling the end of recovery.
The discrepancy between the bone reactions on the external versus the internal surfaces of the osteoderms may stems from the differential vascularization between them, which seems reduced in the external surface. In addition, we suggest that the erosive lesions on the external surface of MLP-PV 98-XII-1-1 and MLP-PV 98-XII-1-2 were much more difficult or impossible to fully recover due to direct and uninterrupted contact with infectious microorganisms that slow down the healing process.
The perforations observed in the carapaces of Panochthus sp. (MLP-PV 16–36) and Hoplophorus euphractus (UFMGMJ.07.1003) represent two potential events of intraspecific fights between glyptodonts. This is the first time the MLP-PV 16–36 specimen has been paleopathologically analyzed, and regarding the perforation in UFMGMJ.07.1003, Paula-Couto proposed a taphonomic explanation for the alteration in this specimen. According to him, the carapace rested on the spine of the caudal tube in the cave environment in which they were found, and this condition caused the perforation on the former (Paula Couto 1957: Fig. 29). However, this explanation does not seem reasonable due to the great resistance of the carapaces to impacts. In addition, the x-ray of UFMGMJ.07.1003 revealed bone response in the edges of the hole, evidencing the ante mortem nature of the lesion (Fig. 7E).
We argue that the alterations in MLP-PV 16–36 and UFMGMJ.07.1003 are related to intraspecific fighting, primarily based on the presence of the lateral figures for the insertion of the horny spines in the genera Hoplophorus and Panochthus and, secondly, on the symmetry between the size of the spines and the morphology of the lesions. Furthermore, in the caudal tube, the distal tubercle on which the spine is inserted has a larger base than the proximal one (Blanco 2009: Fig. 2; Porpino et al. 2010: Fig. 2-g). If the spines were conical, as commonly suggested (Paula Couto 1979; Zurita et al. 2010), and if we maintain the proportions in the size of the tubercles based on the area of insertion, this implies that the distal spine was taller than the proximal one, which would explain why the cranial lesion was larger than the caudal one in MLP-PV 16–36, assuming that they hit each other in inverted positions (Fig. 10) Blanco et al. (2009) pointed out that the region close to the distal tubercle has higher bone density, giving it greater resistance. Furthermore, as it is situated more caudally, the torque of the clubbing movement allows the distal extremities to hit the target more powerfully. In UFMGMJ.07.1003, as there was only one perforation, only the distal spiny seems to have hit the carapace.
Finally, the distance between the two perforations in MLP-PV 16–36 is 18 cm, which is compatible with the distance between the center of the lateral figure (lf1) and the lateral figure (lf2) (= 18–19 cm) of the caudal tube belonging to the same specimen of Panochthus in the exhibition. Assuming that there is body size variation among the individuals of the same species, the difference of 1 cm is negligible in a structure that reaches 80–100 cm or even more, suggesting that the blow was delivered by a glyptodont of similar size.
We can categorize the alterations in MLP-PV 16–36 and UFMGMJ.07.1003 as a comminuted and depressed fracture. Despite observing minor bone response signs close to the lesion through microscopy and x-rays, the osteoderms that formed the perforating site must have been destroyed with the impact, generating a large space among the remaining osteoderms. This prevented the formation of bone callus, in contrast with MLP-PV 98-XII-1-1.
The alteration in the left laterodorsal region of MLP-PV 16–36 resembles two previous records attributed to traumatic injuries in carapaces of glyptodonts. The first, described by Lydekker (1894) in a specimen of Doedicurus clavicaudatus (MLP-PV 16–23), was later attributed to fighting by Alexander et al. (1999) and Fariña (2009). The second is an alteration in Glyptodon reticulatus (MACN-Pv 200) described by Alexander et al. (1999), which we review in this paper (see Glyptodon section). Unfortunately, the MLP-PV 16–23 has undergone some restorations, obscuring the alteration. Still, it was described as a circular-elliptical depressed and fractured area with loss of ornamentation in the right laterodorsal region (Lydekker 1894: plate XXVII; Fariña 2009: Fig. 1), resembling the traumatic injuries we describe here. In MLP-PV 16–36, no fractures were observed, as in the two carapaces of Glyptodon MACN-Pv 200 and MACN-Pv 13776. However, these losses of ornamentation in other points on these carapaces lead us to infer that perhaps combats between glyptodonts did not necessarily result in traumas such as fractures or depressions, but rather wounds on the carapace from blows. We reinforce this by noting that these small lesions without fractures are also very similar to those observed on the left anterolateral region of UFMGMJ.07.1003 (Fig. 7e), which has a fracture but, at the same time, exhibits small lesions not directly associated with the perforation on the right side.
Primitive glyptodonts, such as Propalaeohoplophorus australis, were medium-sized, and their tails were neither armored with spines nor completely ossified like those of Plio-Pleistocene glyptodonts. However, in the diagnosis of P. australis, the author stated: “the tail-sheat in Propalaeohoplophorus is heavy and club-like, tapering but little and ending abruptly, the tube closed by a small irregular plate” (Scott 1903: p. 108). Anatomically, the tail of P. australis would resemble that of Glyptodon clavipes, which also lacked a terminal fully developed caudal tube formed by fused osteoderms typical of the genera such as Panochthus or Hoplophorus. It is accepted that G. clavipes could strike another animal (Fariña 1995), and we provided additional evidence for this in G. munizi, a very similar glyptodont. Therefore, based on Scott´s diagnosis, we hypothesize that even though it did not have a caudal tube, apparently, the tail of P. australis, like that of Glyptodon, would have the attributes to strike another individual in the context of intraspecific fights. In addition, Scott (1903) mentioned a certain uniformity in the morphology of the tail, such as a less mobile end and spiky osteoderms also in other genera within Propalaehoplophorinae (e.g., Asterostemma, Cochlops, and others). Thus, alternatively, the damaged osteoderms observed in MLP-PV 91-II-25-6 may have been generated by disharmonious interaction with other Miocene glyptodonts. In any case, this primitive tail design and the lesions in MLP-PV 91-II-25-6, like those of Plio-Pleistocene glyptodonts, suggest that the fighting behavior was already present in basal glyptodonts during the Miocene.
Pitting
The loss of ornamentation in carapace MACN-Pv 13776 and the breaking of the conical osteoderms on MACN-Pv 12778 must have been caused by the same type of impact, and additionally, we identified two pathological pitting marks on the edge of a single osteoderm which could mimic the marks produced by fracturing. The pitting process creates circumvallate cavities generated horizontally by erosions in the external cortical surface that could reach the trabecular bone (Mathias et al. 2016). Previously, we mentioned that the exposure of osteoderm trabecular bone creates the risk of infection that can trigger the pitting process. In Glyptodon, this mark is similar to the marks observed by Lima and Porpino (2018), suggesting that it is infectious.
Paleopathologies and their relationship with the evolution of armored accessories.
The evolution of a robust carapace and the large size of later glyptodonts suggest a primary defense strategy against predators, owing to their intimidating appearance. This is reinforced by the presence of a carapace with a structure that seems adapted to prevent traumas (Du Plessis et al. 2018) coupled with the possession of well-developed caudal tubes that may have served as potent weapons. However, according to some studies, the caudal tube was not an efficient weapon against fast predators, such as saber-tooth cats (Blanco et al. 2009). On the other hand, due to its weight, it would have been more effective against slower animals, including other glyptodonts. The traumatic alterations we described (see above) support this hypothesis of intraspecific combat. If this latter suggestion holds, we can ask an additional question: what were they fighting for?
Besides the increased size, robust carapace, and well-developed caudal tube, other remarkable characteristics emerged in the more recent lineages of glyptodonts. These include gigantic body sizes, the presence of spines at the end of the caudal tube in certain species, and the development of a carapace composed of even thicker osteoderms capable of withstanding stronger physical impacts (Fariña 1995; Fernicola 2008; Zurita et al. 2010; Zamorano and Fariña 2021). Collectively, these features strongly suggest a combative relationship among glyptodonts, likely evolving through sexual selection.
Although not the primary focus of this study, we put forward two hypotheses about the mechanism that guided the body modifications, assuming they occurred via sexual selection. Firstly, male-male combat for territory and/or partners may have favored an increased body size and the development of attack structures (such as the caudal tube) (Freeman and Herron 2007; Elmen 2008). One strategy adopted by males is to provoke health disorders in their opponents as a consequence of wounds, mutilations, or fractures produced during fights. These injuries are more easily made when pointed structures are present, as in the armors of ankylosaurs (Arbour 2022) or late diverging glyptodonts. Secondly, the complexification of the caudal tube and the appearance of horny spines may have been driven by female preference for attractive-looking males, with features associated with vigor and grandeur that would offer protection to females and their offspring. Nonetheless, to evaluate these hypotheses, further data on tail modifications across various glyptodont species under a phylogenetically informed framework is required, a task beyond the scope of this paper.