OK, coming back to this after I've had a chance to actually read the paper because it's fascinating and I have opinions.
Haolong dongi is a basal Hadrosauroid from the Yixian formation in China. Specifically, it's from the Jianshangou Unit, meaning it shared its environment with Yixian all-stars like Confuciusornis, Sinosauropteryx, and Beipiaosaurus. This specimen is about 2.5 meters long - the size of a large goat - but it's a juvenile, and we don't know how big it would have gotten as an adult. I'm guessing cow-sized.
The specimen preserves extensive preserved integument, covering large swathes of its body. And it is weird. Like the distantly related Kulindadromeus, its tail was covered by large, overlapping feature scales - these were entirely keratinous, lacking underlying osteoderms. The rest of the body was covered in small (~1 mm across) round scales very similar to those found in Psittacosaurus. Interspersed between these scales were keratin spines that seem to be histologically distinct from both scales and protofeathers.
At first glance these spines look a bit like the protofeathers of other ornithischians like Kulindadromeus and Psittacosaurus, but they're not soft filaments - they're genuine spikes, soft at the base and hardened at the tip, chemically identical to the big feature scales of the tail. They come in 2 basic sizes: small (2-7 mm long, 0.1 mm diameter at the base) and large (up to 44.2 mm long, 7.8 mm diameter at the base). The small spines appear to be quite densely packed, growing in the gaps between 1mm scales all over the body, while the larger spines are much less densely distributed over the neck and thorax.
According to the authors, these spines are not "protofeathers" - that is, they don't represent an early stage in feather evolution, and are instead derived features in their own right (I have problems with this conclusion*). They're also not modified scales, since they don't grow out of an enlarged base.
Despite the impeccable quality of preservation, they couldn't find any melanosomes in the spines, so we don't know what colour they were.
The authors speculate about the function of the spines - protection from predators, insulation in cold environments, or sensory bristles. They support the protection hypothesis, and I agree with them for a few reasons: 1, the smaller spines grew between the 1mm scales, so they couldn't have been much more densely packed than, say, a pig's bristles, limiting their use in trapping air for insulation; and 2, the spines are found in parts of the body which would not make sense for a sensory function - the throat, chest, and back.
tl;dr - THIS IS REALLY COOL. Haolong is what I can only describe as a DINOSAURIAN HEDGEHOG. If the spines were used for defense, I wonder if the hollow spaces within the spines could have been used to hold some kind of toxic or irritating fluid (or maybe they "anointed" themselves with toxic saliva or chewed-up poison leaves?) to help deter predators, because the teensy tiny cactus spines seem too wimpy to really deter predators like Yutyrannus or Sinocalliopteryx. Alternately, the smallest spines are about the same size as the glochidia of Prickly Pear cactus - I wonder if they were only loosely attached, and would break off in the skin of any attackers? I wonder if they had microscopic barbs to help anchor them in the skin? I wonder if they had muscles attached and could be raised or lowered as needed? I wonder if the hollow shape and seeming lack of melanosomes made them iridescent, like the bristles of Polychaete worms? PLEASE THIS IS SO COOL I NEED MORE INFORMATION
*Personally, I think the spines are modified protofeathers; maybe I'm misreading it, but the authors seem to dismiss homology with feathers based purely on the fact that the spines are hollow, while "primitive" protofeathers should be solid filaments (as seen in Pterosaurs, Tianyulong, and Kulindadromeus). They use this logic to claim the similarly hollow filaments found in Psittacosaurus are unrelated to feather evolution as well. HOWEVER, this logic depends entirely on the assumption that Ornithischians were incapable of evolving more complex structures, which is completely ridiculous. Idk, the wording in that part of the paper is a little vague - I can't tell if they're claiming the spines are non-homologous to feathers, or if they simply don't represent a primitive condition, but the vibes are making me think the former.
