Introduction

Paleopathological studies are the primary source of information on the diseases affecting extinct animals, but they can also provide essential data on their palaeobiology and behavior (Siegel 1976; Shermis 1983; Beebe and Hulland 1988; Palmqvist et al. 1999; McWhinney et al. 2001; Hanna 2002; Salesa et al. 2006, 2014; Flensborg and Kaufmann 2012; Iurino et al. 2013; Shaw and Ware 2018; Tong et al. 2020; Balisi et al. 2021). Although there are few iconic fossil sites that have yielded dozens of pathological specimens, such as the Pleistocene localities of Rancho La Brea (California, USA), pathological studies on fossil vertebrates does not require large samples, and even an isolated specimen can provide new and interesting information on previously undescribed diseases (Moodie 1930c; Salesa et al. 2014; Hamm et al. 2020; Rabe et al. 2022; Woodruff et al. 2022). The fossil collections from Rancho La Brea include thousands of bones of the saber-toothed felid Smilodon fatalis, and among them is one of the best-known paleopathological collections of a fossil carnivoran, as it gathers a highly diverse sample of bone pathologies rarely seen in other sites (Moodie 1927, 1930a, b, c; Shermis 1983; Binder and Van Valkenburgh 2010; Shaw and Ware 2018). In the present work, we carry out the first in-deep paleopathological analysis of the saber-toothed felid Machairodus aphanistus, thanks to the exceptional sample of this species yielded by the Spanish fossil sites of Batallones-1 and Batallones-3 (Late Miocene, Vallesian, MN10), which includes some specimens showing different pathologies.

The Cerro de los Batallones is a low hill located in the vicinity of Torrejón de Velasco (Madrid, Spain), which has been the focus of mining activities since 1970, mainly due to its rich levels of sepiolite (Morales et al. 2004). Due to this mining activity, a total of nine Late Miocene fossil sites (called Batallones-1, Batallones-2, etc.) were found between 1991 and 2007 (Fig. 1). All of them were formed as irregular cavities due to a geological process called piping, consisting of the erosion of previously deposited sepiolite levels by superficial water flowing along fractures, causing collapses and the development of a karst-like (‘pseudokarst’) system (Pozo et al. 2004; Calvo et al. 2013). At least two of these cavities, Batallones-1 and Batallones-3, acted as natural traps for the carnivorans living in the surroundings of the cavities, and thus, up to 98% of the total number of fossils of large mammals belong to this order. This has provided one of the best-known associations of Late Miocene Carnivora, and has allowed several taxonomic, palaeoecological and functional studies (Abella 2011; Antón et al. 2004; Morales et al. 2004; Monescillo et al. 2014; Peigné et al. 2005, 2008; Salesa et al. 2005, 2006, 2008, 2010, 2019; Siliceo et al. 2014, 2017, 2020). The faunal lists of Batallones-1 and Batallones-3 dominated by the overwhelming presence of two species of saber-toothed felids (subfamily Machairodontinae), the larger, tiger-sized Machairodus aphanistus and the smaller Promegantereon ogygia, of the size of a leopard (Antón et al. 2004; Salesa et al. 2005, 2010; Monescillo et al. 2014; Siliceo et al. 2014). Nevertheless, there are also interesting differences between the carnivoran lists of these 2 sites: they share 8 taxa (the felids Machairodus aphanistus, Promegantereon ogygia, Leptofelis vallesiensis, and Pristifelis attica, the hyaenid Protictitherium crassum, and the amphycionid Magericyon anceps), but there are 2 taxa only recorded at Batallones-1 (the ailurid Simocyon batalleri and the mustelid cf. Martes melibulla), whereas 5 taxa are exclusive of Batallones-3 (the ursid Indarctos arctoides, the amphycionid Ammitocyon kainos, and the mustelids Eomellivora piveteaui, Circamustela peignei and Adroverictis ginsburgi) (Antón et al. 2004; Monescillo et al. 2014; Peigné et al. 2005, 2008; Salesa et al. 2006, 2010, 2019; Abella 2011; Siliceo et al. 2014, 2020; Morales et al. 2021; Valenciano et al. 2015, 2020). Here we carry out a detailed study on the bone injuries that affected the larger species, M. aphanistus, thanks to an exquisitely preserved sample of bones showing this kind of pathologies, and we discuss their relationship to the behavior and degree of sociality of this species of super predator.

Fig. 1
figure 1

Schematic map of the Cerro de los Batallones paleontological complex showing the location of the fossil sites within Spain and the Community of Madrid, and general description of the lithology of the area

Full size image

Materials and methods

The fossils analyzed in this study belong to the species Machairodus aphanistus, and come from the fossil sites of Batallones-1 and Batallones-3, both located at the Cerro de los Batallones (Torrejón de Velasco, Madrid, Spain); they are currently temporarily deposited at the collections of the Museo Nacional de Ciencias Naturales- CSIC (Madrid, Spain). The list of studied pathological material is as follows: B-7284, right calcaneus; BAT-3’09 786b and BAT-3’09 1017, complete mandible fragmented into two portions; and B-6022, left metacarpal III. This material was compared to healthy bones from the same fossil localities. Domingo et al. (2012) already figured the third metacarpal B-6022, but without providing differential diagnoses of the pathology, just indicating that the bone showed an “abnormal regrowth of bone”. That is the reason because we include B-6022 within the studied material in this work.

The studied fossils were photographed with a Canon EOS 50D camera. Photographs were taken in all projections of the bones: cranial, caudal, medial, lateral, and oblique, as well as proximal or distal when necessary. Radiographic images of the fossils for pathological interpretation were carried out in the Diagnostic Imaging Service of the Veterinary Teaching Hospital of the Universidad Complutense de Madrid (Madrid, Spain). Radiographs were taken in the lateromedial and dorsoplantar projections. The radiographic images were later analyzed with the software RadiAnt DICOM viewer ® (Medixant. RadiAnt DICOM Viewer [Software]. Version 2021.1. Jun 27, 2021. URL: https://www.radiantviewer.com).

The anatomical descriptions were carried out following the terminology recommended by the International Committee on Veterinary Anatomical Nomenclature and Illustrated Veterinary Anatomical Nomenclature (Schaller 1996), as well as that used in general anatomical publications such as Barone (2010), Dyce et al. (2012), Nickel et al. (1986) and König et al. (2015). The description and evaluation of the pathological diagnoses were carried out following Craig et al. (2016), Olson et al. (2022) and van Dijk et al. (2007).

Descriptions and diagnostic of the pathological fossils

Calcaneus

Pathological description

The studied pathological calcaneus, B-7284 (Fig. 2a-d), belongs to an adult individual of Machairodus aphanistus. A macroscopically uneven bone growth is developed practically on the entire surface of the fossil, it being more marked on its medial facet (Fig. 2d). Therefore, the tuber calcanei, the coracoid process and the sustentaculum tali are reduced in size, and have lost most of their original morphology, with the two former structures presenting a rough surface due to the large, disordered bone growth developed around them, which shows some irregular cavities or pits (Fig. 2c-d). The actual importance of the lesion can be assessed when comparing the pathological specimen to a normal calcaneus (Fig. 2e-h); it is then evident that there is a strong modification in the morphology of the tuber calcanei in B-7284. The described bone growth suggests an osteitis/osteomyelitis as the main differential diagnosis. This inflammatory response could have its origin in a septicemic process spread through the bloodstream (less probably) or could be secondary to nearby soft tissue infections (muscle, ligaments, tendons, etc.), wound injuries, or external trauma.

Fig. 2
figure 2

Comparison between pathological (a-d) and normal (e-h) calcanei. B-7284, right pathological calcaneus of Machairodus aphanistus from Batallones-1, in: a. dorsal view; b. plantar view; c. lateral view; and d. medial view; B-2584, right normal calcaneus of Machairodus aphanistus from Batallones-1, in e. dorsal view; f. plantar view; g. lateral view; and h. medial view. Abbreviations: cp, coracoid process; st, sustentaculum tali

Full size image

Among the differential diagnoses, bone tumor, chondroma/chondrosarcoma, and osteosarcoma should also be considered, suggested by the disordered bone growth. However, osteosarcomas arise on the distal ends or the metaphysis of long bones such as femur, tibia, humerus or radius, grow aggressively and are locally invasive and can produce severe osteolysis (Craig et al. 2016; Olson et al. 2022). On the other hand, chondromas/chondrosarcomas usually arise on flat bones (head, ribs, pelvis) and are multilobulated (Craig et al. 2016; Olson et al. 2022). A bone callus could be as well considered; it develops as a disorganized network of immature bone formed after a fracture, which functions providing stability to the affected part of the bone (Craig et al. 2016; Olson et al. 2022). Over time, the callus reduces in size and the bone recovers its original shape, although this can take a relatively long period of time, from weeks to several months. Finally, a primary diagnosis of osteoarthritis was ruled out, mainly because the articular surfaces of B-7284 are smooth and did not show any sign of pathological changes, although small osteophytes could be seen (Fig. 2d), but osteoarthritis, together with the main differential diagnosis of osteitis, should not be discarded, considering that external trauma or repeated strains could be probable causes of osteoarthritis (Janssens et al. 2019).

Considering this, as well as the state of development of the callus in the studied individual from Batallones, it is very likely that the death of the animal did not occur much time after the injury.

Radiographic diagnosis

The radiographic images show a marked irregular new bone formation in the plantar and medial aspect of the bone, and a more regular periosteal reaction in its dorsal and lateral margins (Fig. 3a-b). This radiographic pattern is consistent with a moderately aggressive bone lesion, such as osteitis/osteomyelitis, severe osteoarthritis and, less likely, arthritis or bone tumor. Bone callus cannot be rule out, but in that case a concomitant bone infection should be considered to explain de type of periosteal growth.

Fig. 3
figure 3

Radiological images of pathological (a-b) and normal (c) calcanei of Machairodus aphanistus from Batallones-1. B-7284, right pathological specimen in: a. medial view; and b. plantar view; B-2584, right normal specimen, in c. plantar view. Abbreviations: cp, coracoid process; st, sustentaculum tali.

Full size image

Ethological inferences

The calcaneus is the tarsal bone that is most frequently injured in mammals (Lešic et al. 2012), this being more common in those animals that experience high levels of stress in the tarsal articulation, such as cursorial or jumping species. The calcaneal injury can produce an incapacitation of the m. gastrocnemius and the common calcaneal tendon, and, as a consequence, both hyperflexion of the tarsus and hyperextension of the knee, with plantigrade gait, severe lameness, and a marked locomotor disability of the affected limb (Lešic et al. 2012; Craig et al. 2016). In the case of B-7284, the most probable ethological repercussion for the wounded animal would be the development of a disability to carry out activities requiring great physical effort, especially those in which the extremities have a primary role, such as hunting, jumping or running. Such disability to hunt could have led the animal to an early death not long after the injury occurred, since, as mentioned in the pathological description, after the formation of the bone callus, the bone can recover its original shape over time, and, as can be seen in the studied specimen, the alterations are still evident and the bone had lost its normal morphology. It should be noted that in most calcaneal injuries, including fractures, the common calcaneal tendon exerts a great traction force on the fractured or injured parts, which makes it difficult to resolve the fracture (Perry et al. 2017). For example, tearing of ligamentous attachments often induces local periostitis and the development of exostoses or osteophytes (Craig et al. 2016). In the case of small exostoses, they may be completely resorbed if the stimulus disappears, but larger ones may be converted from woven to lamellar bone and persist indefinitely (Craig et al. 2016). Exostoses often remain clinically inapparent but may interfere with the function of adjacent structures, such as tendons or ligaments, and there is evidence suggesting that traumatic injury to the periosteum may predispose to bacterial periostitis at the site, even if the overlying skin is intact.

A wide range of possibilities should be considered as a possible etiology for the calcaneal injury seen in the specimen from Batallones; among them, fights with other predators, injuries during hunting, repeated trauma from different activities, and bone weakness secondary to metabolic or infectious pathologies. Unfortunately, it is very difficult to elucidate which one could have caused the pathology.

Metacarpal III

Pathological description

The studied metacarpal III of Machairodus aphanistus from Batallones-1, B-6022 (Fig. 4) shows a diffuse periosteal and endosteal new bone growth that uniformly covers both corpus and basis, and a narrowing of the medullar cavity. Thus, the lateral edge of the bone is more prominent and irregular on the lateral edge (Fig. 4b), and protrudes from the bone surface, increasing its diameter and forming several structures with a beaked morphology. Based on the nature of these lesions, the most probable diagnosis is a osteosclerosis, this is, a diffusely thickening of the cortical bone with increased bone density (van Haul et al., 2001), but an hyperostosis, or thickening of the periosteum throughout the entire bone (Brickley and Ives 2008, Craig et al. 2016) and occasionally reduction of the medullary cavity (Touraine et al. 2013) cannot be ruled out because of the beaked morphology of the bone.

Fig. 4
figure 4

B-6022, right pathological Mc III of Machairodus aphanistus from Batallones-1, in: a. lateral view; b. dorsal view; c. medial view; and d. palmar view. Abbreviations: ba, basis; bo, body; h, head; II, articular facet for the Mc II; IV, articular facet for the Mc IV; mg, articular facet for the magnum. Scale bar equals 5 cm

Full size image

Osteosclerosis is a generic designation with several different entities which are diagnosed in the basis of radiographic findings, among which osteopetrosis and hyperostosis corticalis generalisata (hereditary diseases) together with some acquired diseases, are considered (Ihde et al. 2011). Osteopetrosis is a bone dysplasia which main feature is an increase bone density (van Haul et al. 2001) due to accumulation of primary spongiosa in marrow cavities and a defective osteoclastic bone resorption (Craig et al. 2016). It is a group of rare disorders with several different subtypes, most of which with genetic basis, that has been diagnosed in humans, cattle, horses, and dogs (Craig et al. 2016). However, this disease usually appears in stillborn or young animals and the bones are fragile and prone to fractures (Craig et al. 2016). In adult cats, along with other adult animals, a disease with similar morphologic and radiographic features, has been diagnosed associated to viral infections and other systemic pathologies (Craig et al. 2016; Wright et al. 2003) but because of the dissimilarities with osteopetrosis, the name of acquire osteopetrosis or generalized osteosclerosis is a better choice.

So, in this case, though osteopetrosis should be considered as a differential diagnosis, the main option is osteosclerosis, or osteopetrosis-like osteosclerosis, or acquired osteopetrosis. As mentioned above, in extant cats, acquired osteopetrosis or osteosclerosis has been associated with feline leukemia virus (FeLV) infections, and secondary to lymphoma (or other neoplasias), myeloproliferative disease, or immune diseases (such systemic lupus erythematosus) (Wright et al. 2003; Craig et al. 2016). However, the rarity of osteosclerosis in domestic animals makes it hard to understand its pathogenesis or to identify the possible causes.

Hyperostosis should also be considered as a differential diagnosis. The excessive periosteal bone formation can occur because of different insults such a subperiosteal hemorrhage, trauma, metabolic and infectious diseases, vitamin and mineral imbalance or neoplasia, but, sometimes, there is a marked hyperostosis without a direct cause (Craig et al. 2016). So, many differential diagnoses should also be considered, being trauma or inflammatory diseases the most probable. Toxic bone diseases such as osteofluorosis (fluoride toxicity) should be included; this disease occurs due to an excessive intake of fluoride from subsurface waters rich in phosphates, though herbivores are much more susceptible (Craig et al. 2016). As said above, idiopathic hyperostosis can occur, so hypertrophic osteopathy should also be considered.

Radiographic diagnosis

Radiographic images of the metacarpal III (Fig. 5) show a diffuse thickening of the body cortices with significant reduction of the medullary cavity and loss of normal trabecular bone pattern. The endosteal margins are delimited by a sclerotic fine line in the distal half of the body. This radiographic appearance suggests osteosclerosis. Though, in cases of hyperostosis a thickening of the periosteum and reduction of the medullary cavity can be also seen (Touraine et al. 2013). These radiographic findings could also be consistent with osteopetrosis, a rare form of bone dysplasia characterized by osteosclerosis, described, although not well characterized, in canids and felids (Stander and Cassel 2016).

Fig. 5
figure 5

Radiological images of B-6022, right pathological Mc III of Machairodus aphanistus from Batallones-1, in: palmar (left) and medial (right) views. Abbreviation: mc, medullary cavity; Scale bar equals 5 cm

Full size image

Ethological inferences

Although in many cases, new bone growth is asymptomatic, in other cases it can also affect adjacent structures, such as muscles, tendons, nerves and blood vessels, producing inflammatory processes, swelling, pain and claudication of the limb (Voss 2009; Harasen and Little 2012). As a result of the described injury in the metacarpal III, the animal could have suffered pain in its forelimb and hand, with the consequent functional repercussions on the locomotor and hunting capacities, primarily if the hyperostosis was related to trauma, inflammation or neoplasia. Saber-toothed felids subdued their prey with their powerful foreclaws, so they needed a great force to keep their prey immobilized while struggling trying to escape (Akersten 1985; Rawn-Schatzinger 1992; Lewis 1997; Turner and Antón 1997; Turner et al. 2011). In this hunting scenario, the hands play an important role, and thus, the pain produced by the pathology in the left metacarpal III could have resulted in an incapacity for using at least this hand, which affected both locomotion and hunting. Nevertheless, the animal survived long enough after the first stages of the injury, until it was trapped and died in Batallones-1, so the observed pathology could have been moderately painful, producing less severe consequences and allowing the animal to be able to hunt and feed. In any case, in the absence of more data on the rest of the skeleton, it is impossible to assess which of the two situations occurred.

Mandible

Pathological description

The studied mandible of Machairodus aphanistus from Batallones-3, is broken in two portions (labelled BAT-3’09 786b and BAT-3’09 1017) (Fig. 6) and belongs to an adult individual as shown by the completely erupted m1. Only two teeth were preserved in the specimen: the right lower canine and the left m1, the latter severely broken at the level of the paraconid. Located in the left mandibular body, extending from the p4 (which is not present) to the m1, a focal, nodular, well-demarcated lesion is observed. The lesion produced an anfractuous cavity of rounded morphology (Fig. 6d), divided into two parts of variable size, (Fig. 6c, g), which are consistent with bone necrosis or resorption, as well as new bone formation of irregular shape and periosteal proliferation, and thus the most probable diagnosis is an inflammatory osteomyelitis. The cavities could be filled with inflammatory exudate (most probably pus) and necrotic bone, which suggests the formation of an abscess. One of the most common sequelae of abscesses is the occurrence of pathologic fractures (Craig et al. 2016), as occurred in the studied fossil, leaving the left hemimandible completely divided into two pieces.

Fig. 6
figure 6

BAT-3’09 1017 and BAT-3’09 786b, pathological mandible of Machairodus aphanistus from Batallones-3; BAT-3’09 1017 in: a. right lateral view; b. occlusal view; c. caudal view of the abscess surface; d. left lateral view; BAT-3’09 786b in: e. right medial view; f. left lateral view; g. rostral view of the abscess surface. Abbreviations: ap, angular process; c, right lower canine; m1, lower m1 (partially broken); mef, mental foramina; mf, mandibular foramen; cop, coronoid process; cp, condylar process. Scale bar equals 5 cm

Full size image

Infectious pyogenic osteomyelitis in the mandible and maxilla of carnivorans occurs, normally, via local extension of severe bacterial periodontitis (Craig et al. 2016), local puncture wounds, as previously described in two Pleistocene American lions (Beebe and Holland 1988), or dental infection after dental rupture. This seems to be the case with the mandible of M. aphanistus from Batallones, whose broken m1 probably constituted the entrance to suppurative bacteria.

As a differential diagnosis, the possibility of actinomycosis should be considered, although it is uncommon in carnivorans (Craig et al. 2016). Actinomycosis is a mandibular osteomyelitis caused by a bacterium of the genus Actinomyces (Harz 1877), very common in ruminants, which acquire the infection when the microorganism, which is a normal inhabitant of the oral microbiota, invades the bone through the surrounding soft tissues because of mucosal damage caused by sharp objects or erupting teeth (Smith 2002; Park et al. 2005; Gensa 2018). This disease is characterized by the production of granulomas or pyogranulomas in the bone tissue, mainly in the mandible, accompanied by new bone formation (reactive bone) and bone resorption (Smith et al. 2011; Valour et al. 2014) that offers a “honeycomb” appearance on the cut (Craig et al. 2016).

A second differential diagnosis to be considered is a fracture, although the absence of other evidence, together with the failure of bone remodeling to return the mandibular body to its normal condition suggests a different etiology. And finally, the development of a bone tumor should be considered, although the absence of bone destruction or alteration of the surrounding mandibular bone makes this diagnosis very unlikely.

Radiographic diagnosis

The radiographic image shows a fractured focal lesion located in the central part of the left mandible body. The ends of the fragments appear widened by a quite extensive, regular and homogenous new bone formation, with slightly irregular and concave margins, that could limit a cavitary lesion in origin (Fig. 7). In that case, the fracture could be secondary to osteolytic changes in the center of the lesion (pathologic fracture), although a fracture related to the fossilization process cannot be ruled out. These radiographic findings are consistent with a mandibular abscess as main differential diagnosis. Another less probable differential diagnosis would be a low-grade bone tumor.

Fig. 7
figure 7

Radiological image of BAT-3’09 1017 and BAT-3’09 786b, the two pieces of the pathological mandible of Machairodus aphanistus from Batallones-3, placed in their probable position in the living animal. Scale bar equals 5 cm

Full size image

Ethological inferences

The most probable etiology for the development of the abscess in the mandible is a dental fistula (although other structures adjacent to the teeth may also be involved, such as the periodontium), which can lead to the destruction of the mandibular body (Lombardero et al. 2021). In the studied specimen, the most probable origin of the mandibular abscess, based on the fracture of the m1, is a pulpal infection secondary to a crown fracture. This fracture could be the consequence of a puncture wound with penetration of a foreign body, or a complication of severe periodontal disease. Sabre-toothed felids killed their prey with a precise bite on its throat, avoiding any contact to vertebrae or other bones in order to prevent canine breakage (Antón and Galobart 1999; Antón et al. 2004, 2020; Emerson and Radinsky 1980; Salesa et al. 2005, 2006; Turner and Antón 1997). Nevertheless, even considering this avoidance behavior, during hunting, prey usually defends itself by kicking or goring, and this can cause severe lesions to the predator. There are few examples of fossil mandibles of carnivorans showing similar pathologies, probably because this is a very important part of the predator anatomy, and any lesion affecting it, could seriously compromise the survival of the animal. A study by Iurino et al. (2013) on a mandible of the canid Cuon alpinus from the Upper Pleistocene of San Isidero (Italy), which presented chronic periodontitis in the left maxilla at the level of the molar and premolar teeth, concluded that the trauma did not occur during the predation, when canines and incisors are primarily used, but probably during feeding on a hard object, as canids usually chew hard items like bones or cartilages (Kaunda and Skinner 2003; Jhala and Moehlman 2004; Lanszki et al. 2009). In our specimen of M. aphanistus, it is difficult to infer the origin of the observed lesion but, considering the mentioned bone avoidance behavior of saber-toothed felids, it is highly improbable that the injury occurred when feeding, it being more likely as a consequence of a hunting accident. In any case, in this kind of injury, once the bacterial infection has entered through the pulp and tooth cavities, it spread to the surrounding oral tissues, producing more local inflammation, swelling and pain in the area (Bellows et al. 2019). With these symptoms, the hunting capacities of the animal would have been greatly diminished, which would have caused the weakness and even death of the individual in a relatively short period of time. In fact, this condition probably encouraged this individual to enter the Batallones cavity searching for food, then being trapped into the hole. In a study by Beebe and Hulland (1988) on two hemimandibles of Pleistocene American lion (Panthera atrox) which showed chronic periodontitis and extensive sclerosal osteomyelitis, these authors reached similar conclusions. These individuals, similar to the studied specimen of M. aphanistus, showed a pathological process that produced an increase of bone mass with evident remodeling and widening in the mandibular ramus, which suggested that the most probable etiology was bone necrosis, localized osteomyelitis, or vascular trauma, although this latter option was described as being less probable (Beebe and Hulland 1988). These authors also discussed that this was a process lasting several months, and the mandible infections could have started by a local puncture wound, perhaps a tooth penetration in a fight with another lion, or a porcupine quill during an attempt to hunt one of these animals. The absence of an obvious fistula and the large amount of new bone formation that occurred in the healing process suggest that the infection was primarily pyogenic. On the other hand, the initial infection could have been controlled by the immune system, leaving necrotic bone as the only sign of the disease (Beebe and Hulland 1988).

The pathological mandible of M. aphanistus studied here would have suffered the lesion for several months, as suggested by the bone changes observed in the specimen, the increase in bone mass, the remodeling process, and the widening of the left mandibular body. In domestic cats, the fractures in the mandibular body take more time to heal than those in other mandibular regions, with an average time lapse of 10 weeks, and maximum of around 16 weeks (Glyde and Lidbetter 2003; Lombardero et al. 2021), and this could have been the case of the individual of M. aphanistus when the infection produced after the dental fistula evolved in a complete separation of the hemimandible in two pieces, moreover considering that the inflammatory process continued, and there was no possibility for the hemimandible to heal on its own. The constant pain that the injury probably caused to this animal prevented it from developing a normal hunting activity, moreover considering the specialized technique of saber-toothed felids, in which the mandible was used as an anchor whilst a strong flexion movement of the head applied the mortal bite to the throat of prey (Emerson and Radinsky 1980; Turner and Antón 1997; Antón and Galobart 1999; Antón et al. 2004; Salesa et al. 2005, 2006). Despite this, the individual of M. aphanistus survived for several weeks until it finally entered into the Batallones-3 cavity, probably to feed on the carcass of a previously trapped animal. The study by Beebe and Hulland (1988) suggested that the immune system of the two pathological specimens of P. atrox was very capable of controlling the injury in some way, leaving the bone altered, but without severe repercussions, and thus the animals could have survived for a much longer time. Nevertheless, in the case of our specimen of M. aphanistus, the extension and gravity of the injury probably resulted not only in the fracture of the mandible, but also in septicemia that had led to the death of the animal in case it had not been trapped in the Batallones trap.

Paleobiological inferences from the pathologies present in the fossils of Machairodus aphanistus from the Batallones sites

Fossil bones with pathologies are an excellent source of information on the diseases suffered by extinct predators, and although provide direct data on the life story of single individuals, they also allow to infer several aspects on their locomotor adaptations, hunting behavior, ecology and physiology (Moodie 1927, 1930a, b; Shermis 1983; Beebe and Hulland 1988; Palmqvist et al. 1999; McWhinney et al. 2001; Hanna 2002; Iurino et al. 2013; Salesa et al. 2014; Tong et al. 2020; Balisi et al. 2021). With carnivorans being relatively scarce in the typical fossil mammalian localities (around 7% of the total sample) (White et al. 1984; Antón and Morales 2000), the study of the paleopathologies in this group has been always hindered by this low relative abundance. For that reason, those few fossil sites with a high percentage of carnivorans constitute an excellent source for the study of the paleopathology of this group. This is the case of the two fossil localities from Cerro de los Batallones studied here (Batallones-1 and Batallones-3), where the percentage of carnivorans reaches the 98–99% (Antón and Morales, 2000; Salesa et al. 2006; Domingo et al. 2011; Monescillo et al. 2014; Martín-Perea et al. 2021); in the carnivoran sample yielded by these sites, M. aphanistus is represented by thousands of cranial and postcranial elements, thus increasing the possibility of finding pathological examples, as shown in this work.

The different pathologies observed in the individuals of M. aphanistus from Batallones probably had strong influence on many aspects of their behavior. Nevertheless, to what extent these pathologies impaired their overall health and how the specimens survived until the moment of their entrapment into the cavities of Batallones is difficult to assess. To find this out, it would be necessary to have more data on the ethology and social structure of this large felid, in order to infer if an injured individual could have survived feeding on the carcasses hunted by other members of the group, as is the case of lions, or, on the contrary, the solitary nature of their behavior would have led to the death of fatally injured animals. Palmqvist et al. (1999), for example, carried out a study on a population of Canis (Xenocyon) falconeri, a hypercarnivorous canid from the lower Pleistocene fossil site of Venta Micena (Spain), which included an individual with different dental anomalies and pathologies that survived at least six months from the injury, according to this authors; this suggests that the survival of this animal was possible due to the development of a cooperative behavior, similar to that of modern canids such as the African wild dogs (Lycaon pictus) or the wolf (Canis lupus). In the case of felids, which are essentially solitary animals, but with notable exceptions, this kind of inference must be taken with more caution. Most of the extant species of felids, except the lion, Panthera leo (and tigers -P. tigris- and cheetahs -Acinonyx jubatus- under certain conditions), have a mainly solitary life, forming couples or small groups during the breeding season for short periods of time (Novak and Paradiso 1983; Bailey 1993; Alderton 1998; Bothma and Walker 1999). Also, wild felids are territorial animals that exhibit a high degree of competition for resources, both interspecific and intraspecific, killing adults and young of other species of carnivorans (Alderton 1998; Bailey 1993; Bothma and Walker 1999; Morse 1974; Turner and Antón 1997). In addition, the existence of a high degree of sexual dimorphism in some felid species, such as leopards (Panthera pardus), tigers and lions, produces frequent aggressive interactions between males for access to females and for keeping the territory free of other males (Bailey 1993; Short and Balaban 1994; Turner and Antón 1997; Alderton 1998; Bothma and Walker 1999). Nevertheless, there are also differences between extant felids, with jaguars (Panthera onca) being more tolerant to the presence of other adult individuals in their territory than leopards and lions (Rabinowitz and Nottingham 1986; Seymour 1989). This is remarkable for the study of the effects of pathologies in fossil felids, as important differences have been described in the surviving rate to bone injuries between those species showing some degree of sociality (lions) or tolerance to the presence of other adults (jaguars), and those being intolerant to the presence of other individuals, such as leopards (Bailey 1993). Thus, although temporarily injured leopards can eventually feed on carrion (Houston 1979; Bailey 1993; Daniel 1996; Bothma and Walker 1999), those with severe injuries die before healing, as they are not tolerated within the territories of other individuals, whereas lions in similar circumstances survive due to their social system (Schaller 1972; Hanby and Bygott 1979; Bailey 1993). Considering these observations, Salesa et al. (2006) studied a pathological individual of the leopard-sized saber-toothed felid Promegantereon ogygia from Batallones-1, which showed the four left metatarsals broken and healed in life, but fused incorrectly, which produced a severe limping that prevented the animal from normal walking, running and hunting activities. This allowed an interpretation of this fossil felid as having a behavior similar to that of extant jaguars, and thus an injured individual could have been feeding on the carcasses left by other adult individuals until its entrapment into the Batallones-1 cavity (Salesa et al. 2006). Other paleopathological studies on saber-toothed felid fossils have also suggested the existence of some degree of social behavior (Shaw and Ware 2018; Rabe et al. 2022; Balisi et al. 2021), and recently, comparative genomic analyses on Homotherium humerus from Dominion Creek site (near Dawson City, Yukon Territory, USA) found signatures of positive selection in several genes, including those related to well-developed social behavior (Barnett et al. 2020).

The pathologies observed in the individuals of M. aphanistus studied here were in some cases very severe, especially those affecting the calcaneus (B-7284) and the broken mandible (BAT-3’09 786b and BAT-3’09 1017). Among domestic cats, although tarsal injuries are relatively common (Schmökel et al. 1994), severe injuries of the calcaneus are rare (Carmichael and Marshall 2012; Cantatore and Clements 2015) and tend to be developed as a consequence of high mechanical stress or external trauma (Schmökel et al. 1994; Carmichael and Marshall 2012). There are few data on the percentage of this injury in wild felids, but these same causes could be attributed to the presence of this pathology, and even a higher frequency should be expected given the stressful life of wild species in comparison to that of domestic cats. For example, fractures due to external trauma can be linked to accidents during hunting, jumping or running, or even during intraspecific fighting, whereas those originated as consequence of high mechanical stress are classified into two main categories: (1) fractures caused by fatigue: this are due to repeated stresses applied to the bone, which produces an accumulation of microdamage due to excessive cyclic loading beyond the threshold for repair (Taylor 1997; Muir et al. 1999; Cantatore and Clements 2015; Perry et al. 2017); and (2) insufficiency fractures, originated when normal stresses are repetitively applied to an abnormal bone (Daffner and Pavlov 1992). There are several reports of stress fractures in domestic cats, mostly affecting the ribs, patellae, proximal tibia and calcaneus (Arnbjerg and Bindseil 1994; Hardie et al. 1998; Langley-Hobbs 2009; Langley-Hobbs et al. 2009; Cantatore and Clements 2015), but it is difficult to assess if the pathology observed in the calcaneus of M. aphanistus from Batallones was produced by stress or by an external trauma. In wild felids, injuries in the calcaneus, including fractures caused by external trauma can be related to running, fighting, jumping or hunting accidents, but it is so difficult to interpret the exact parameters involved in that kind of injury, that its presence provides few data on the predator ecology. On the contrary, those injuries derived from high mechanical stress can be more informative, as when related to repeated stresses applied to the bone, they allow inferences on the locomotor or hunting behavior (Salesa et al. 2014), as well as on the biomechanics of the limbs and other body parts. The calcaneus forms part of an important anatomical region subject to high tensile forces during locomotion (Evans 1993; Barone 2010), which has been suggested as a possible reason for severe injuries due to fatigue in cats (Cantatore and Clements 2015). In any case, the calcaneus B-7284 shows a disorganized periosteal growth, and a severe malformation. As it has been observed in domestic cats, such calcaneal pathology can produce malfunction of the common calcaneal tendon attached onto the tuber calcanei, and those affecting the distal end tend to result in plantar instability (Piermattei et al. 2006; Longley et al. 2016). This is a cause of non-weight bearing lameness and plantigrade stance (Longley et al. 2016), which would definitely affect the locomotor and hunting activities of the animal.

But even more severe is the case of the individual with the injured mandible (BAT-3’09 786b and BAT-3’09 1017), a pathology that is very informative to evaluate the impact of this lesion on the biting and feeding capacity of the animal. As discussed above, considering the specialized hunting technique of saber-toothed felids, it is very improbable that this injured individual could capture and kill prey. Despite this, this individual of M. aphanistus did survive for several weeks until it entered the Batallones-3 cavity, probably to feed on the carcass of a previously trapped animal. The question about how this individual survived until its entrapment poses the possibility that it had access to carcasses hunted by other individuals. Does that imply that M. aphanistus lived in groups as lions do? This does not have an easy answer, as social systems in felids are a rare behavior, developed in response of very specific environmental and ecological requirements, such as body size, prey range and density, or the defense of cubs and territory in relatively open environments (Leyhausen 1979; Packer 1986; Caro 1989, 1994; Turner and Antón 1997; Bothma and Walker 1999; Yamaguchi et al. 2004). On the other side, it has been observed that, although consumption of carrion is habitual among injured and healthy leopards (Houston 1979; Bailey 1993; Daniel 1996; Bothma and Walker 1999) those individuals showing severe injuries die before healing, whereas lions in similar circumstances show a high percentage of survival thanks to their social system (Bailey 1993). The presence of saber-toothed felids in the Batallones traps (Antón et al. 2004; Salesa et al. 2006; Monescillo et al. 2014; Siliceo et al. 2014) supports scavenger behavior, but carrion is not a constant and abundant resource in extant ecosystems, so it seems insufficient to explain the survival of the severely injured individual of M. aphanistus in Batallones-3. Thus, it could be possible that this species developed some kind of social system, although it is difficult to determine if it was similar to that of lions, with groups of females defended by a male coalition, or more similar to that of cheetahs, in which males form groups that control solitary females (Caro 1994; Alderton 1998; Bothma and Walker 1999; Durant 2000). The Batallones sites have yielded thousands of fossils of saber-toothed felids, most of them belonging to adult individuals, the fossils of juvenile being very rare (Antón et al. 2004; Salesa et al. 2006). But even this absence of juvenile individuals is not indicative of a specific social structure, as among extant felids, cubs and young do not patrol or hunt with the adults (Bailey 1993; Antón and Morales 2000; Packer et al. 2001; Bradshaw 2016); thus, the virtual absence of cubs and juveniles of M. aphanistus in Batallones does not provide any data on possible sociality. Another variable to be considered is the sexual dimorphism index, which is very high in M. aphanistus, similar to that of extant lions and leopards (Salesa et al. 2006). Generally, those carnivoran species in which a male defends a group of females tend to be more dimorphic than those with monogamous pairs or groups of males and females (Gittleman and Van Valkenburgh 1997; Weckerly 1998; Van Valkenburgh and Sacco 2002). Given this, M. aphanistus would fit within the first group, and thus males would be very aggressive to the presence of other males in their territories, although this does not exclude the possibility that males of this saber-toothed felid formed coalitions of two or more adult individuals to defend a large area including females territories, which in this case would be more or less solitary animals.

Conclusions

In conclusion, the studied sample of pathologies of M. aphanistus from Batallones-1 and Batallones-3 supports the existence in this species of a certain degree of sociality, which favored the survival of injured specimens, even those with severe pathologies such as a broken and infected hemimandible. It is difficult, however, to determine precisely which kind of social structure existed, as extant felids show a wide range of social interactions depending on several variables related to predators density, type of habitat, and prey availability (Leyhausen 1979; Packer 1986; Caro 1989, 1994; Turner and Antón 1997; Bothma and Walker 1999; Yamaguchi et al. 2004), and these data are difficult to infer from most of the fossil sites. In any case, the Batallones fauna suggests that the sites were formed in a relatively structured habitat, this is, composed by wooded areas with enough vegetal cover to allow the coexistence of several species of carnivorans, next to more open, grassy patches (Salesa et al. 2006). This could indicate that M. aphanistus did not live in prides, as this kind of social structure is observed only among lions, which occupy almost exclusively savannas and open landscapes (Leyhausen 1979; Caro 1989, 1994); it is interesting that tigers can also form groups, but this has been observed only in populations occupying open environments (Alderton 1998; Bothma and Walker 1999). In the case of M. aphanistus, the degree of social interactions between adult individuals could have consisted of males forming coalitions of two or three individuals to defend a large area including the territories of multiple females (Fig. 8), which would explain the presence of severely injured individuals in the Batallones sample that survived for a relatively long time before being trap into the cavities.

Fig. 8
figure 8

Reconstruction of a hunting scene in the Cerro de los Batallones during the Late Miocene: two adult males of Machairodus aphanistus have subdued and killed a three-toe horse of the genus Hipparion. Artwork by M. Antón

Full size image