Sunday, April 7, 2019

Titanosaur osteoderms: characteristics

Now that we've had a look at the background and history of study for titanosaur osteoderms, we can get into the nitty-gritty of their anatomy and where they were positioned on the body. It turns out there's still quite a lot we don't know about these structures.


Classifying titanosaur osteoderms has proven difficult because they vary quite a bit in size, shape, and microanatomy (Cerda et al. 2015). Each specimen is kind of its own thing. This doesn't mean that no one has tried; rather, several schemes have been proposed. Huene (1929) may not have known what Loricosaurus really was, but he split its osteoderms into "crested" and "uncrested" plates, plus irregular ossifications. Le Loeuff et al. (1994) and Le Loeuff (1995) divided the four osteoderms of Ampelosaurus into spines, bulbs, and scutes or plates, with the bulb type eventually becoming the bulb and root of later researchers (to prevent confusion from unclear word usage, I will refer to these as bulb-and-root). D'Emic et al. (2009) proposed four types: ellipsoidal, keeled (a keel on the internal surface, the surface facing into the body), cylindrical (flat internal and external surfaces), and mosaic (including ossicles). Vidal et al. (2014), using only complete specimens, recognized ossicles (mm-scale, distinct in histology), scutes, and bulb-and-root osteoderms. Most recently, Cerda et al. (2015) played it simple, splitting them into ossicles (mm-scale) and plates (cm-scale). While I certainly sympathize with this instinct, I'm going to go with the Vidal et al. scheme, which seems to best capture the basic distinction between osteoderms that are mostly scute and osteoderms with a lot going on beneath the scute. However, the D'Emic et al. system is also frequently used.


The first thing to establish is how we can tell a given specimen came from a titanosaur, and not a croc or ankylosaur. Generally, titanosaur osteoderms are more roughly textured and irregular than ankylosaur or croc osteoderms; they tend to be lumps or discs of fairly rough bone. Sanz and Buscalioni (1987) differentiated titanosaur osteoderms from ankylosaur and croc osteoderms by the presence of a well-developed cingulum (a surrounding belt or ridge, a feature later found to not be diagnostic; Marinho and Candeiro 2005), nodular surface texture, and the appearance of internal ducts, while Le Loeuff et al. (1994) found the shape and texture to be distinct.

Under the Vidal et al. scheme, scutes are the classic Saltasaurus-type osteoderms. The D'Emic et al. system uses "keeled" for the same type, with the name referring to the distinctive longitudinal keel on the ventral surface (=facing into the body). Not all have keels, though, nor do all have cinguli. They can also be described as similar to the bulb part of bulb-and-root osteoderms (Vidal et al. 2014). The basic shape is oval to circular, with a smooth and frequently keeled internal surface and a roughly textured external surface, sometimes bearing a crest, blunt point, or depression (Bonaparte and Powell 1980; Powell 2003). The shape and texture of the scutes make them vaguely resemble coarse oatmeal cookies, and they're similar in size, too: published examples range from 10 cm long (3.9 in) (Bonaparte and Powell 1980) to 18 cm long (7.1 inches) (Salgado et al. 2005). Scute osteoderms are comparatively rare and only definitely known from Madagascar and South America, which is the source of the scutes of Loricosaurus, Neuquensaurus, and Saltasaurus (Vidal et al. 2014; the Australian report omitted as discussed in the previous post).

A scute from Saltasaurus (Cerda and Powell 2010:Figure 1). From the top down, the views are of the external surface, ridge-on, and ridge-side (the usual assumption is that lengthwise features like the ridge indicate the osteoderm was oriented with the ridge lengthwise on the body). CC-BY-4.0.

A bulb-and-root osteoderm, when complete, tends to look something like a slug wearing a hat, the "slug" being the root and the "hat" being the bulb, the external expression of the osteoderm. (More formally, they're described as teardrop-shaped or amgydaloidal, almond-shaped.) The bulb is set off from the root by a cingulum, and can be concave, flat, or convex (bluntly pointed) (Vidal et al. 2014).

Here are two examples of bulb-and-root osteoderms with convex bulbs, from Vidal et al. (2014:figure 3). The A "slug" appears to be wearing Devils Tower, while the B "slug" has a more conservative beret. Scale bar is 10 cm. CC-BY-4.0.

Most of the "ellipsoid" and some of the "cylinders" of the D'Emic et al. (2009) classification scheme are partial to complete bulb-and-root osteoderms, and an incomplete bulb-and-root osteoderm can look like a scute osteoderm (Vidal et al. 2014). Vidal et al. (2014) regarded the bulb-and-root form as the more primitive form of titanosaur osteoderm, and more widespread as well, with examples from Africa, Europe, Indo-Pakistan, Madagascar, and South America.

Reconstructing partial bulb-and-root osteoderms, from Vidal et al. (2014:figure 8). Scale bar is 10 cm. CC-BY-4.0.

Bulb-and-root osteoderms can be quite large: a specimen from the Upper Cretaceous of Patagonia is 37.6 cm long (14.8 inches) (Cerda et al. 2015); Vidal et al. (2014) reported a couple in excess of 50 cm long (20 inches) from the Lo Hueco site in Spain; and the large Rapetosaurus krausei osteoderm described in Curry Rogers et al. (2011) is 57.2 cm long (22.5 in). Interestingly, bulb size seems to be correlated to body size, with consistent bulb length/femur length ratios in several taxa (Carrano and D'Emic 2015).

The two largest osteoderms from Vidal et al. (2014:figure 4), both root-and-bulb osteoderms with concave bulbs. Scale bar is 10 cm. CC-BY-4.0.

Some bulb-and-root osteoderms have large internal hollows, an odd feature for putative armor. The champion example is the large Rapetosaurus krausei specimen cited above (Curry Rogers et al. 2011). Other osteoderms with significant internal cavities include specimens reported for Alamosaurus sanjuanensis (Carrano and D'Emic 2015) and from Lo Hueco (Vidal et al. 2017). Voids may have begun under the bulb and spread into the root (Vidal et al. 2017). As to why the osteoderms should have voids... well, check the next osteoderm post.

An osteoderm with a lot of nothing (the red blob), from Vidal et al. (2017:figure 4). Scale bar is 10 cm. CC-BY-4.0.

In general, titanosaur scutes and bulb-and-root osteoderms are well-vascularized (Salgado 2003) and probably had a nonbony covering, perhaps a keratinous material (Depéret 1896; Bonaparte and Powell 1980; Salgado 2003; Vidal et al. 2017) or epidermal scales (Curry Rogers et al. 2011). Pathologies have been found in some osteoderms, including irregular cavities and abnormal overgrowths, which may be the results of parasites, infections (Cerda et al. 2015), and bites (Marinho and Iori 2011; Cerda et al. 2015).

Ossicles are so far are only known from Saltasaurus and perhaps "Lametasaurus". Some reported "ossicles" appear to be the remains of larger crushed hollow osteoderms (Curry Rogers et al. 2011). It is possible that ossicles were more widely distributed than is known; the poor ossicle record may be due to taphonomic biases (e.g., getting washed away while larger bones remain) and/or collecting biases (e.g., being unattractive to collectors) (D'Emic et al. 2009; Cerda et al. 2015). Saltasaurus ossicles are about 7 to 10 mm across (about 0.25 to 0.4 inches), with smooth flat external surfaces, rounded internal faces with woven texture, and subspherical to lenticular shapes. They are packed into mosaics of 18 to 25 ossicles per 10 square cm (less than two square inches) (Cerda and Powell 2010). Unlike the larger osteoderms, they are poorly vascularized (Cerda et al. 2015). The mosaics resemble patterns of scales seen in titanosaur embryos (Chiappe et al. 1998; Cerda and Powell 2010; Vidal et al. 2014).

Seems like a logical place to reuse this illustration of ossicles, Figure 3 from Cerda and Powell (2010). With so few samples, there aren't a lot of choices to pick from. CC-BY-4.0.


The locations of osteoderms on titanosaurs is not known, although some educated guesses have been made. There are several problems facing the researcher. First off, because we're dealing with titanosaurs, much larger animals than ankylosaurs, we're already up against a relative taphonomic bias against articulated specimens. Next, there is no evidence that any titanosaur osteoderms (not counting the little ossicles) touched or imbricated with each other (Curry Rogers et al. 2011; Cerda et al. 2015). In addition, dinosaur osteoderms in general don't leave marks on underlying bones, so they can't be placed that way. The sad fact is even if a skeleton was found with osteoderms in life position, it would be very difficult for us to *know* that; how could we be sure there was no displacement?

One further complication is that the much-vaunted armored sauropods really had very little armor (Dodson et al. 1998; D'Emic et al. 2009; Curry Rogers et al. 2011; Vidal et al. 2014, 2017; Cerda et al. 2015). As of about a decade ago, only about 90 osteoderms had been reported worldwide (D'Emic et al. 2009; Curry Rogers et al. 2011). Even the armor of Saltasaurus loricatus is represented by a grand total of eight scutes (or seven at one point in Bonaparte and Powell 1980, down to six in Powell 2003) and four collections of ossicles (Bonaparte and Powell 1980), all from a quarry that produced at least four non-juvenile Saltasaurus loricatus and apparently at least one non-juvenile Neuquensaurus australis (D'Emic and Wilson 2011). We aren't talking rows and rows of scutes and plates, but a handful of large elements per sauropod at most.

Bonaparte and Powell (1980) suggested that osteoderms were located on the back and sides. Sanz and Buscalioni (1987) put them in the sacro-pelvic region and possibly the tail, but Sanz et al. (1999) backed off. Le Loeuff et al. (1994), dealing with the relatively heavily armored Ampelosaurus atacis, put flat bulbs on the torso and pelvic region and convex bulbs over the shoulders. González Riga (2003) suggested that the osteoderms of Mendozasaurus may have been located on the tail and/or the sacral and lumbar region. Salgado (2003) seems to have been the first to make the reasonable propositions that symmetric osteoderms were located along the midline of the animal and asymmetric osteoderms were displaced from the midline. Cerda et al. (2015) supported these ideas and further suggested that pieces with a strong curve were positioned over narrower parts of the body (neck or tail). These strongly curved osteoderms also sometimes come with spine-like cones; Cerda et al. (2015) proposed that pointed or keeled osteoderms may have been on the tail, for striking. Vidal et al. (2014) hypothesized that asymmetric bulb-and-root osteoderms could have been positioned in two rows similar to stegosaur plates.

Ossicles are a different case and may well have covered a significant part of the body of the few (one?) titanosaurs they are known from. Bonaparte and Powell (1980) reported that an ossicle accumulation had been found in the posterior-dorsal region of an articulated Saltasaurus pelvis. As noted, the ossicle mosaics resemble patterns of scales seen in titanosaur embryos (Chiappe et al. 1998; Cerda and Powell 2010; Vidal et al. 2014), but it is not known if ossicles and scutes, which would have grown after hatching, would have corresponded to the scale pattern, nor exactly where the skin impressions were on the body (Vidal et al. 2014).


Bonaparte, J. F., and J. E. Powell. 1980. A continental assemblage of tetrapods from the Upper Cretaceous beds of El Brete, northwestern Argentina (Sauropoda-Coelurosauria-Carnosauria-Aves). Mémoires de la Société Géologique de France 139:19–28.

Carrano, M. T., and M. D. D'Emic. 2015. Osteoderms of the titanosaur sauropod dinosaur Alamosaurus sanjuanensis Gilmore, 1922. Journal of Vertebrate Paleontology 35(1):e901334. doi:10.1080/02724634.2014.901334.

Cerda, I. A., and J. E. Powell. 2010. Dermal armor histology of Saltasaurus loricatus, an Upper Cretaceous sauropod dinosaur from northwest Argentina. Acta Palaeontologica Polonica 55:389–398.

Cerda, I. A., R. A. García, J. E. Powell, and O. Lopez. 2015. Morphology, microanatomy, and histology of titanosaur (Dinosauria, Sauropoda) osteoderms from the Upper Cretaceous of Patagonia. Journal of Vertebrate Paleontology 35(1):e905791. doi:10.1080/02724634.2014.905791.

Chiappe, L. M., R. A. Coria, L. Dingus, F. Jackson, A. Chinsamy, and M. Fox. 1998. Sauropod embryos from the Late Cretaceous of Patagonia. Nature 396:258–261.

Curry Rogers, K., M. D. D’Emic, M. Vickaryous, and A. Cagan. 2011. Sauropod dinosaur osteoderms from the Late Cretaceous of Madagascar. Nature Communications 2:564. doi:10.1038/ ncomms1578.

D'Emic, M. D., and J. A. Wilson. 2011. New remains attributable to the holotype of the sauropod dinosaur Neuquensaurus australis, with implications for saltasaurine systematics. Acta Paleontologica Polonica 56(1):61–73.

D'Emic, M. D., J. A. Wilson, and S. Chatterjee. 2009. The titanosaur (Dinosauria: Sauropoda) osteoderm record: review and first definitive specimen from India. Journal of Vertebrate Paleontology 29(1):165–177.

Depéret, C. 1896. Note sur le dinosauriens sauropodes et théropodes du Crétacé Supérieur de Madagascar. Bulletin de la Société Géologique de France 24:176–194. [translated]

Díez Díaz, V., X. Pereda Suberbiola, and J. L. Sanz. 2013. Appendicular skeleton and dermal armour of the Late Cretaceous titanosaur Lirainosaurus astibiae (Dinosauria: Sauropoda) from Spain. Palaeontologia Electronica 16.2.19A. doi:10.26879/350.

Dodson, P., D. W. Krause, C. A. Forster, S. D. Sampson, and F. Ravoavy. 1998. Titanosaurid (Sauropoda) osteoderms from the Late Cretaceous of Madagascar. Journal of Vertebrate Paleontology 18:563–568.

González Riga, B. J. 2003. A new titanosaur (Dinosauria, Sauropoda) from the Upper Cretaceous of Mendoza Province, Argentina. Ameghiniana 40:155–172.

Huene, F. von. 1929. Los Saurisquios y Ornitisquios del Cretáceo Argentino. Anales del Museo de La Plata 3:1–196.

Le Loeuff, J. 1995. Ampelosaurus atacis (nov. gen., nov. sp.), un nouveau Titanosauridae (Dinosauria, Sauropoda) du Crétacé supérieur de la Haute Vallée de l'Aude (France). Comptes Rendus de l'Académie des Sciences Paris 321 (series IIa):693–699.

Le Loeuff, J., E. Buffetaut, L. Cavin, M. Martin, V. Martin, and H. Tong. 1994. An armoured titanosaurid sauropod from the Late Cretaceous of southern France and the occurrence of osteoderms in the Titanosauridae. Gaia 10:155–159.

Marinho, T. S., and C. R. A. Candeiro. 2005. Titanosaur (Dinosauria: Sauropoda) osteoderms from the Maastrichtian of Uberaba, Minas Gerais State, Brazil. Gondwana Research 8:473–477.

Marinho, T. S., and F. V. Iori. 2011. A large titanosaur (Dinosauria, Sauropoda) osteoderm with possible bite marks from Ibirá, São Paulo State, Brazil. Pages 369–379 in I. D. Souza Carvalho, N. K. Srivastava, O. Strohschoen, Jr., and C. C. Lana, editors. Paleontologia: Cenários de Vida 4. Editorial Interciência, Rio de Janeiro.

Powell, J. E. 2003. Revision of South American titanosaurid dinosaurs: palaeobiological, palaeobiogeographical, and phylogenetic aspects. Records of the Queen Victoria Museum 111.

Salgado, L. 2003. Considerations on the bony plates assigned to titanosaurs (Dinosauria, Sauropoda). Ameghiniana 40:441–456.

Salgado, L., S. Apesteguía, and S. E. Heredia. 2005. A new specimen of Neuquensaurus australis, a Late Cretaceous saltasaurine titanosaur from North Patagonia. Journal of Vertebrate Paleontology 25(3):623–634.

Sanz, J. L., and A. D. Buscalioni. 1987. New evidence of armored dinosaurs in the Upper Cretaceous of Spain. Pages 199–204 in P. M. Currie and E. H. Koster, editors. 4th Symposium of Mesozoic Terrestrial Ecosystems, Drumheller, Alberta, 10–14 August 1987. Short Papers. Royal Tyrrell Museum, Paleontology, Drumheller, Alberta.

Sanz, J. L., J. E. Powell, J. Le Loeuff, R. Martinez, and X. Pereda-Suberbiola. 1999. Sauropod remains from the Upper Cretaceous of Laño (northcentral Spain). Titanosaur phylogenetic relationships. Estudios del Museo de Cienceias Naturales de Alava 14(especial 1):235–255.

Vidal, D., F. Ortega, and J. L. Sanz. 2014. Titanosaur osteoderms from the Upper Cretaceous of Lo Hueco (Spain) and their implications on the armor of Laurasian titanosaurs. PLoS ONE 9(8):e102488. doi:10.1371/journal.pone.0102488.

Vidal, D., F. Ortega, F. Gascó, A. Serrano-Martínez, and J. Luis Sanz. 2017. The internal anatomy of titanosaur osteoderms from the Upper Cretaceous of Spain is compatible with a role in oogenesis. Scientific Reports 7, article number 42035. doi:10.1038/srep42035.


  1. Any word on the supposed spiked osteoderms of Spinophorosaurus? Well knowing that one abstract hints they are no such thing...

    1. Hmm... well, it's not too likely that basal sauropod tail osteoderms have much to do with titanosaur osteoderms, given the apparent lack of continuity. The elements pictured by Remes et al. in the description of Spinophorosaurus (Fig. 4A-C) don't look much at all like any kind of titanosaur osteoderm. On the other hand, if the abstract (Vidal et al. 2015, "The Specialized Tail of Spinophorosaurus nigerensis..." right?) is correct that they are not osteoderms, a lot of bones in a skeleton can be reduced to triangular shapes.