Controversial ceratopsids revisited: woolly Pachyrhinosaurus and scavenging Styracosaurus

Spurned on by a print request, I've spent free time this week revising two images of ceratopsids which may be familiar to long-term readers: my woolly Pachyrhinosaurus perotorum and scavenging Styracosaurus albertensis. The former is now just over two years old, and the latter a whopping eight years old - wow, have I really been messing about with internet palaeoart for that long?

Maastrichtian Alaska was quite chilly, but woolly Pachyrhinosaurus perotorum doesn't care. See this post for the original image and exploration of the concept shown here. Prints are available.

Because I appreciate some folks are fond of my original paintings, I haven't deviated too far from the original compositions and instead just added more detail, tweaked colour values and tidied up some sketchy areas. I'm very conscious of not 'pulling a Lucas' on my old work. Most importantly, the science has been improved/corrected: the cranial morphology of Pachyrhinosaurus perotorum is now correct to that species (like a doofus, I based the morphology in the original image on a different Pachyrhinosaurus species) and the tyrannosaur in the scavenging scene is appropriately filamentous. Thanks to Darren Naish, Zachary Miller and Christian Kammerer for discussions of Styracosaurus horn shape.

Have these depictions have been supported or refuted by any new discoveries? As far as I'm aware, skin impressions still remain elusive for Pachyrhinosaurus, although new data has emerged on the facial integument of juveniles (Fiorillo and Tykoski 2013). The 2014 discovery of Kulinadromeus and its assortment of filament-like scales, true filaments and other integumentary oddities (Godefroit et al. 2014) might indirectly add credence to the idea of shaggy ceratopsids, however. Along with Psittacosaurus and Tianyulong, Kulindadromeus shows that the evolution of ornithischian integument was complex, that single animals can bear a suite of different integument types, and that the assumption of dinosaur skin being ancestrally scaly is uncertain. The weird scales in (unpublished) skin impressions of Triceratops are further evidence that 'one skin fits all' approaches to reconstructing these animals are likely flawed, and that even clades with relatively limited anatomical disparity - like ceratopsids - had diverse integuments. Thus, the idea that some members of Dinosauria may have looked very different to our traditional interpretations is being strengthened by genuine data, and shaggy arctic ceratopsids remain a fun extrapolation of that concept. For further discussion on these points, check out my discussion of version one of the woollysaur painting.

The Campanian centrosaurine Styracosaurus albertensis scavenges the remains of a tyrannosaurid. He was going for warpaint on his face, but he ended up at 'Tonto'. For fun, the original 2007 version can be seen here. Prints are available.

What of scavenging ceratopsids, as in the reworked 2007 image of a tyrannosaurid-eating Styracosaurus? Ceratopsid omnivory has yet to be explored in the technical literature and, to my knowledge, remains best represented by a short paragraph in Paul (1991). More recently, Mallon and Anderson (2013) provided reasoning for why ceratopsids were not predatory animals, although their discussion seems to consider 'carnivory' synonymous with 'predation': opportunistic scavenging or omnivory are not explored. This leaves most discussion of ceratopsid scavenging online, and several famousdenizens of the online palaeontological community seem to support it. Back in 2007 I wrote a long essay substantiating the idea. That essay is no longer online*, but the argument is pretty straightforward:

*After eight years, I figured it's time to archive my old Flickr stream. The bulk of the content there is not representative of modern science or a good representation of my work, so it's been taken offline. I won't pretend I'm not a bit sad to do so, but there's obviously reason for bringing internet searches to my best, most recent work, not images I created when first learning how to paint.

1. As is well-known, a number of modern herbivores eat animal remains on occasion. This may reflect nutrient stress (thought to explain carrion use by hippos, which is not as common as 'common knowledge' might suggest) or else a method of supplementing a mineral-deficient diet (as in deer, cows, giraffes and a host of other hoofed mammals - Hutson et al. 2013). Remarkably, some cases of hippo carnivory involve the hippos killing animals first, and they will also scare other carnivores from kills to obtain carcass access (Dudley 1998). Of further interest is that entire herds of hippos will chew on carcasses when available - these are not the acts of rogue, aggressive or aberrant individuals (Dudley 1998). Note that studies on the carnivorous tendancies of generally herbivorous animals are in their infancy, and it may be that this behaviour is more common and opportunistic than we currently realise.
2. Other species, such as pigs, ingest animal matter as part of their normal diets. Studies on some pigs suggest 28% of their diet is derived from animals, either being invertebrates or carrion (e.g. Thomson, and Challies 1988). There is no reason to think that large extinct animals were incapable of comparable omnivory, but we restrict most discussion of it to smaller dinosaurs and pterosaurs. We can predict that such animals should have jaws mostly adapted for herbivory (e.g. teeth suited to browsing and grazing, long 'cheek' toothrows, vertically displaced jaw joints etc.) but would also have some means to process animal remains (e.g. crushing teeth to break bones, caniform teeth or sharp beaks for ripping meat etc.).
3. Ceratopsid jaws certainly belonged to primarily herbivorous species capable of chewing their food, but their approach to herbivory was unusual. Their teeth and jaws, unlike other herbivorous dinosaurs and mammals, were incapable of grinding plant matter. Instead, they sliced food into pieces, their teeth sliding vertically past one other like scissors. Ceratopsid beaks are also unusually deep and narrow compared to other dinosaurian herbivores, and recall the beaks of parrots in many respects. The beaks of these birds are famously powerful, enabling their owners to access a range of nuts, seeds and animal matter (e.g. Greene 1999). The diet of of ceratopsids has been questioned by palaeontologists because chopping plant matter is not common among modern herbivores. To the contrary, most food slicers are carnivores - meat is easier to chop and slice into easily digested chunks than it is to grind into a paste. One sensible suggestion is that ceratopsids ingested particularly fibrous, woody plant matter (see Mallon and Anderson 2013 and references therein). We might imagine them devastating Cretaceous shrubs, removing entire chunks of tree - leaves, branches and bark - with each bite, or overturning plants with their huge heads to access their roots and tubers. However, it is odd that their jaws aren't more convergent with those of other herbivores, as grinding mechanisms have developed so many times in multiple tetrapod lineage and might be considered optimal for breaking down plant matter. So, maybe ceratopsid jaws were used for more than simply eating plants, and their shearing teeth and hooked beaks are the traits of omnivory we mentioned above, equally capable of slicing plants and animal remains. Opening carcasses, snapping smaller bones and slicing meat was almost certainly possible with their jaws and beaks, and we might imagine ceratopids as Mesozoic variants of pigs: largely herbivorous species with opportunistic carnivorous tendencies, and certainly capable of competing with strict carnivores for carcass access. The possibility that they could occasionally kill other animals for food, as demonstrated by the aforementioned hippos, is not unreasonable.

Back in 2007 I mentioned a possible smoking gun for this idea - a rumoured Psittacosaurus specimen with bony gut content. Since then, it's become apparent that that specimen either doesn't exist, has disappeared or has otherwise been forgotten about - it's best to consider that an unsubstantiated rumour for now. Despite this, I still think the concept of ceratopsian omnivory has legs: maybe a technical paper on the topic would be worthwhile.