Pathological changes in bone tissue are the result of imbalances in the normal equilibrium of bone resorption and formation or growth-related disorders, being the causal agents related to mechanical stress, changes in blood supply, inflammation of soft tissues, infectious diseases, hormonal, nutritional and metabolic upsets, and tumours, amongst others (White & Folkens, 2005). Skeletal diseases may be expressed as abnormal bone formation, abnormal bone destruction, abnormal bone density, abnormal bone size, and abnormal bone shape, and each of these bone expressions may occur as the only manifestation of disease in a skeleton or as a combination of one or more of the other expressions (Ortner, 2003). Buikstra & Ubelaker (1994) recognized that one of the rules to assist the diagnosis of a pathological condition is the evidence of an inflammatory reaction or process of healing, which indicates that the animal survived for some time after suffering an injury. Together with the advent of new tools and techniques, the ability to make accurate diagnoses of diseases in fossils has increased greatly, sometimes with the same methods and the same degree of certainty as in living patients (Martin & Rothschild, 1989). Thus, the correct identification of a palaeopathology can be helpful to our understanding of the origin, distribution, and evolution of a disease, and at the same time, help to understand a number of issues such as predator−prey relationships, intraspecific interactions, and disease or trauma experienced by an animal (e.g. Rothschild & Martin, 1993; Rothschild et al., 1998; Sawyer & Erickson, 1998; Tanke & Currie, 1998; Rothschild, Witzke & Hershkovitz, 1999; Hanna, 2002; Wolff, Fowler & Bonde, 2007). However, when studying diseases in fossil remains one of the major problems that we find is distinguishing them from biostratinomic and diagenetic damage (post-mortem abnormal modifications resulting from mechanical, environmental, biological, and/or chemical processes). Some post-mortem changes can mimic pathological lesions and lead to misinterpretation. Biogenic changes indicate post-mortem processes in burial sites, and provide evidence of trophic and behavioural interactions within ancient communities (Rogers, 1992; Jacobsen, 1998; Roberts, Rogers & Foreman, 2007). Insect-generated bone modifications are particularly useful in the understanding of trophic and behavioural interactions because of the extreme sensitivity of insects to local conditions and to the use of a wide array of biological materials for feeding, reproduction, and shelter (Roberts et al., 2007).
Despite an increase in studies regarding fossil vertebrates (Kellner & Campos, 1999), studies of diseases in Brazilian fossils are not very common. In the case of fossil reptiles, pathological features were mentioned only by Kellner & Tomida (2000), who observed broken ribs and a probable infection in the skull of a pterosaur, and Azevedo, Henriques & Gomide (1995), who reported rehealed ribs in a titanosaurid dinosaur from the Bauru Group. Recently several new crocodyliforms have been collected from Brazilian deposits (e.g. Carvalho, Vasconcellos & Tavares, 2007; Barbosa, Kellner & Viana, 2008; Kellner et al., 2009; Campos et al., 2011) but in none were pathological features described. Here we report some abnormal features observed in the type specimen of Stratiotosuchus maxhechtiCampos et al. 2001 (DGM 1477-R), a baurusuchid mesoeucrocodylian from the Upper Cretaceous, Bauru Basin, Brazil. They are of pathological nature, the first record of these features in a crocodyliform from Brazil. Furthermore, burrows perforating some elements were also observed and are regarded as insect borings. All these abnormalities are here described and discussed.
DGM, Departamento de Geologia e Mineralogia, Rio de Janeiro, Brazil; UFRJ DG, Universidade Federal do Rio de Janeiro, Departamento de Geologia.
The fossil remains of S. maxhechti reported here were collected in the Irapuru city outcrops (Campos et al., 2001), which are located in the west of São Paulo State, near to the junction between the SP-501 and SP-294 roads, in the south-west part of Bauru Basin. This basin is a large cratonic depression that developed during the Late Cretaceous in the central south-eastern portion of the South American Platform (Fernandes & Coimbra, 1996). It represents the Cretaceous coverage overlying the spills of the Serra Geral Formation with an area of about 370 000 km2 and a maximum preserved thickness of 300 m (Riccomini, 1997; Fernandes & Coimbra, 2000; Fernandes, 2004).
Several stratigraphical divisions of the Bauru Basin have been proposed for the São Paulo State (Soares et al., 1980; Fernandes & Coimbra, 1996; Fernandes & Coimbra, 2000) and even today there is no consensus on this subject. The classical division was proposed by Soares et al. (1980) and includes the Caiuá, Santo Anastácio, Adamantina, and Marília Formations. Fernandes & Coimbra (2000) redefined the stratigraphical division of the Bauru Basin and recognized the follow units in the west of São Paulo State: Santo Anastácio and Rio Paraná Formations (both included in the Caiuá Group) and the Bauru Group, including the Marília Formation and a reinterpretation and subdivision of the former Adamantina Formation in the Vale do Rio do Peixe, Araçatuba, São José do Rio Preto, and Presidente Prudente Formations.
Unfortunately, many fossils from the Bauru Basin were found by chance, with no precise stratigraphical control, as it is the case for the type specimen of S. maxhechti. However, outcrops of the municipality of Irapuru are considered to be present in the Presidente Prudente Formation, owing to their geographical location in the deposits of the Bauru Basin. The lithological characteristics of the matrix in which the specimen was found concur with this stratigraphical positioning.
The Presidente Prudente Formation occurs in portions of the valleys of the Peixe and Paranapanema rivers, in the region between Presidente Prudente and Adamantina cities, in the west of São Paulo State (Fernandes & Coimbra, 2000). Its thickness is about 50 m and lays gradually in interdigital mode over the Vale do Rio do Peixe Formation (Fernandes, 2004). The matrix where the fossil was included consists of a massive, fine-grained, friable sandstone with beige to light-brown coloration, bearing sparse and subrounded intraclasts of mudstone (pellets), which have a slightly darker brown coloration and millimetric diameters, with a maximum diameter of 2 cm, without carbonatic cementation. The depositional environment of the Presidente Prudente Formation is interpreted as fluvial meandering, of shallow channels and low sinuosity, with facies of alluvial plains (Fernandes & Coimbra, 2000; Fernandes, 2004).
The holotype of S. maxhechti (DGM 1477-R) is an articulated and almost complete skeleton, including the skull, with no significant distortion or compression (Campos et al., 2001; Riff, 2007). There is no record of the relative position and attitude of the skeleton on the field, but a block containing the anterior dorsal vertebrae and the right scapular girdle and member, articulated and complete (including the phalanges and carpal elements), was not prepared until the beginning of the detailed description of the skeleton (Riff, 2007). Analysing this material, we inferred that no significant transport occurred and that desiccation was taking place at the time of fossilization, as the anterior member was preserved with a strong flexion.
The observed abnormalities on the anterior and posterior limbs of the DGM 1477-R S. maxhechti specimen deviate from the osteological pattern expected for this taxon and for that reason were analysed in order to determine the cause of these changes.
The roughened surfaces of right metacarpal V suggest that the lesions were in the process of being remodelled (with absorption of the bone callus). Likewise, the bone callus in left metatarsals I and II indicate that the lesions were in the process of healing. In addition, these features were not found in any other element of S. maxhechti. These reactionary bone growths allow us to refute the possibility that the damage resulted from scavenging or another post-mortem modification, being considered instead to be a result of disease or trauma.
According to the set of characteristics found in the right metacarpal V of S. maxhechti (a radiolucent line crossing the bone, formation of irregular trabeculae, an exuberant bone callus, and the displacement of the proximal end of the bone), the differential diagnosis includes fracture, post-traumatic infection, and tumour.
Regarding bone healing in extant reptiles, little information is available; however, it appears that it occurs at a significantly slower rate when compared with birds and mammals (Mader, 1996). Bone healing in reptiles takes about 6–30 months but 4–8 weeks and 2–6 weeks in mammals and birds, respectively. There are several factors that can affect bone healing, including fracture type, patient age, environmental temperature, and nutritional status (Mitchell, 2002). Moreover, unlike mammals, which routinely produce a significant periosteal reaction in response to a bone fracture, most reptiles produce only a subtle periosteal reaction (Mitchell, 2002). Nowadays it is common to find individuals with traumatic injuries resulting from aggressive intraspecific behaviour in disputes for territory and mates. This kind of behaviour in extant crocodilian populations has been documented in the literature through direct observations based on captured animals. According to Webb & Messel (1977), several injuries observed in Crocodylus porosus originated when attacking prey, when being attacked by predators, or when involved in conflicts related to social behaviour and territoriality. Likewise, Webb & Manolis (1983) attributed the scars and injuries found in Crocodylus johnstoni to the same causes as for C. porosus. However, the right metacarpal V of the S. maxhechti presents the opposite of the pattern expected for healing in reptiles: large areas of bone destruction and new bone formation. For this reason, it is more likely that the extensive bone neoformation of the right metacarpal V was triggered from another pathological condition, causing an intense lytic and sclerotic process.
Some forms of infection and tumour affect the underlying bone structure in addition to forming a callus (Schulp et al., 2004). An infectious process can be started after a bone fracture and consequently the original shape of the bone can be strongly altered under the callus as seen in the right metacarpal V of the holotype of S. maxhechti. In addition, Schulp et al. (2004) also describe the development of pus drains, similar to the lesions observed in the metacarpal studied here. Nevertheless, sequestrum, a diagnostic characteristic of infection, is absent.
It is interesting to note that tumours (of lytic, mixed, or even sclerotic origin) may result in pathological fractures and thus the great bone formation and the fracture in right metacarpal V of S. maxhechti could be explained as the result of a cancer. Bone cancer can result in the formation of irregular trabeculae (Rothschild et al., 1999), a feature that is also found in this bone. Moreover, cancer precludes bone healing in most pathological fracture situations, which could explain the unhealed status of the fracture. According to macroscopic, microscopic, and radiological observations, these two diseases could be interpreted as post-traumatic infection and tumour, reflecting the injuries found in the right metacarpal V of S. maxhechti.
The set of bone modifications observed in both metatarsals is consistent with stress fracture. Stress fractures are alterations in bone that take place in response to unusual, submaximal, repeated stress forces (Rothschild & Martin, 1993). Stress fractures of the lower extremities most commonly involve the tibia and metatarsal bones, and although several factors appear to contribute to the development of stress fractures they generally occur as a result of a repetitive use injury that exceeds the intrinsic ability of the bone to repair itself. (Sanderlin & Raspa, 2003). Prolonged, unconditioned, strenuous activity produces progressive microfractures and promotes resorption (accelerated remodelling) of bone, which precipitates such stress fractures (Rothschild & Martin, 2006). Animal studies have demonstrated that bone subjected to repetitive cyclical loading also develops these microfractures (Reeser, 2009).
Rothschild (1988) noticed stress fracture in ceratopsian phalanges and more recently in theropods (Allosaurus and Ceratosaurus) and sauropods (Diplodocus, Apatosaurus, Camarasaurus, and Nurosaurus) (Rothschild & Martin, 2006). This kind of pathology has also been found in tyrannosaurids (Rothschild, Tanke & Ford, 2001).
The CT scan images also suggest the diagnosis of stress fracture because there is no line of fracture (which would be expected for a complete fracture), even in the inner bone layer. In the same way, the bone callus does not involve the whole bone and there is no angulation or displacement of it. Their aetiology in man typically relates to marching, heavy lifting, or prolonged standing (Rothschild, 1988). Although the S. maxhechti lesion may be related to long periods of standing, a more likely scenario is sudden exertions related to their aggressive behaviours.
Although not directly related to the animal's death, the lesions found in right metacarpal V, and in both metatarsals I and II of S. maxhechti, may have contributed to it as the ability to obtain food tends to be reduced; the animal tends to be more susceptible to other diseases and complications, and also has an increased risk of being predated.
The other three elements in DGM 1477-R, the left ulna and both tibiae, right and left, also present some abnormalities. However, although some features in these specimens suggest pathologies, such anomalies may have another origin. The differential diagnosis in this case includes abscesses, multiple myelomas, gout, erosive arthritis, and taphonomical disorders.
The presence of abscesses is a condition that is related to some types of disease such as infections and tumours. They are characterized as a lytic lesion, usually elliptical, that is orientated along the long axis of the bone and the cavity margins are typically sclerotic and sharply defined, occasionally associated with widening of the bone itself (Rothschild & Martin, 1993). However, the elements analysed here do not present any evidence of reactive sclerosis, a typical condition of an abscess.
Multiple myelomas are characterized by several ‘punched out’ holes in the bone, decreased bone density on X-rays, and by lesions with effaced/erased trabeculae (Rothschild & Martin, 2006). However, these last two characteristics cited are not observed in the specimens studied here.
Gout may have a punched-out appearance that is accentuated by a sclerotic margin and, in addition radiology is very helpful in the diagnosis of this condition because it may well demonstrate sclerotic margins, the presence of a Martel hook or, rarely, interosseous calcification (Waldron, 2009), but none of these features are observed in the specimen described here.
The bone changes in erosive arthritis are mostly confined to the hands and in radiography these changes appear as a mixture of proliferation (osteophytes) and bone erosions (Waldron, 2009), features which are not found in the left ulna or either tibiae of S. maxhechti.
Based on the observations above, the differential diagnosis of the abnormalities in the left ulna and both tibiae allowed us to discard the possibility of abscesses, multiple myelomas, gout, and erosive arthritis. Thus, the most likely explanation, supported by elimination of the others previously cited, is that the borings were generated by some kind of taphonomical disorder. The borings, however, cannot be assigned either to tooth marks, as there are no signs of crushing of bone near the perforation, or to plant roots, because they generally leave more irregular borings and cause bone fragmentation as they grow. The alterations found here can be regarded as boring marks, which are bioerosion structures produced mainly by insects. Unfortunately, the borings were not detected before the bones were extracted from the sediment, and thus the position of such borings in the outcrop context remains uncertain.
Bone modification provides a critical insight into the post-mortem history of vertebrate fossil assemblages (Roberts et al., 2007). These modifications can be caused by mechanical, environmental, chemical, and biological factors and have been well documented in bone fossils as trampling marks, tooth marks, scratch marks, weathering, breakage, and bioerosion structures (Behrensmeyer, 1978; Rogers, 1992; Jacobsen, 1998; Rogers, Krause & Curry Rogers, 2003; Fejfar & Kaiser, 2005). Bioerosion structures result from biochemical or mechanical excavation produced by an organism in consolidated substrate (rocks, trunks, shells, or bones) (Fernandes et al., 2002). The organisms producing these structures in bones may be other vertebrates or even invertebrates. Nevertheless, the form and depth of the borings found in DGM 1477-R indicate that the organisms responsible for these modifications were invertebrates, the scavengers and butchers being excluded as possible producers. Insect generators of bone modifications are useful in providing evidence of behaviour and trophic interactions within ancient communities because they are extremely sensitive to local conditions and they utilize a wide array of biological materials for feeding, reproduction, and shelter (Roberts et al., 2007).
Rogers (1992) identified the occurrence of macro-borings in bones of an undescribed species of Prosaurolophus (Hadrosauridae dinosaur) from the Upper Cretaceous of Montana, USA, suggesting the activity of carrion beetles. The borings in these bones are solitary, nontapering shafts, filled with sediment identical to the host matrix of the bone bed and with no indication of surface modification. Schwanke & Kellner (1999) registered the first occurrence of insect boring marks in Triassic bones from Brazil. These marks were recognized in a left humerus of a dicynodont that shows several cylindrical perforations with a diameter of about 6 mm. They present no preferential orientation, but each one tends to be retilineous and does not change in diameter. The borings were attributed to dermestid pupation chambers (Schwanke & Kellner, 1999). Roberts et al. (2007) reported some types of borings in dinosaur bones from continental deposits of the Upper Cretaceous Maevarano Formation of Madagascar and the Upper Cretaceous Kaiparowits Formation of southern Utah, USA, and attributed these borings to necrophagous or osteophagous insects that were presumed to have been constructed for purposes of reproduction and feeding. Dominato et al. (2009) recognized several perforations, represented by hollow oval-shaped structures (without filling), excavated in the spongy part of the bone, in vertebrae of Pleistocene mastodons of Minas Gerais State, Brazil, that were identified as being related to the ichnofossil Cubiculum ornatus, a ichnospecies associated with a dermestid (Coleoptera).
The characteristics of borings found in DGM 1477-R are very similar to those reported previously and suggest that the organism producers of the borings was a carrion insect, most likely associated to dermestids. This is supported by the aspect of the borings and by the sediment that filled the inner tunnel on the right tibia indicating that the borings were created before the fossilization of the bones. Both the borings and the innermost part of the bone are filled by sediment and therefore were not permineralized as were the other bones that still had some organic matter. The occurrence of dermestids indicates extended subaerial exposure under warm conditions and a palaeoclimate susceptible to desiccation (Rogers, 1992; Martin & West, 1995). The presence of such borings in the holotype of S. maxhechti extends the still very scarce record of insect bone borings in Brazilian fossils, and represents the first record in vertebrate fossils of the Presidente Prudente Formation.