Monday, May 18, 2009

Horns & Spikes, Part 1: Postorbital Horns

Ah, ceratopsian dinosaurs. They are quickly becoming my favorite terrible lizards, and in this series of posts (Horns & Spikes), I intend to discuss the various horns, spikes, and generally pointy structures that decorate the typical ceratopsid skull. For convenience's sake, I will here only cover the Ceratopsidae--with one or two invocations of most basal taxa where proper. The series will consist of four parts: Postorbital Horns, Nasal Horns & Bosses, Epoccipitals, and Jugal "Horns." Granted, that last section will be pretty short. I hope you enjoy the series! And here's part one:

Postorbital Horns

The postorbital horns are among the most obvious features of the ceratopsid skull: their great length in most chasmosaurines is impressive, and their absence in most centrosaurines is striking. Postorbital horns are basal for Ceratopsidae, occuring in the family's outgroup taxa, Zuniceratops and Turanoceratops, and having been inhereted by both the Chasmosaurinae and Centrosaurinae. Basal members of both groups have elongate postorbital horns. Surprisingly, the basalmost centrosaurines, Albertaceratops and an undescribed Utah taxon, have more exaggerated postorbital horns than basal chasmosaurines! The postorbital horns of Chasmosaurus (C. belli and C. kaiseni) are relatively short--but thick--while those of Albertaceratops are surprisingly long and erect.

As centrosaurines evolved, however, they seem to have lost interest in retaining their postorbital horns. Styracosaurus and Centrosaurus both retain short, rather stubby postorbitals early in life, but these are reabsorbed and shrink away to nothingness during adulthood. A similar process occurs in Einiosaurus. Pachyrhinosaurines, however, make use of their postorbitals differently: the postorbital horns of juvenile animals are small and laterally compressed. With growth, however, the postorbitals grow not upwards, but laterally, forming eventually forming wide bosses that nearly meet in the middle of the skull. Oddly, old adults still reabsorb postorbital tissue, and the surface becomes pitted or, in places, hollow. In some specimens of Pachyrhinosaurus lakustai, both the postorbital and nasal bosses are actually concave due to reabsorption!

A sampling of ceratopsid postorbital horncores. From top to bottom, left to right: "Monoclonius flexus" (AMNH 5239), Styracosaurus albertensis (CMN 344), Albertaceratops nesmoi (TMP 2001.26.1), Achelousaurus horneri (MOR 485), Chasmosaurus kaiseni (AMNH 5401), Pentaceratops sternbergi (OMNH 10165), Pachyrhinosaurus lakustai (TMP 1989.55.1234), Triceratops horridus (SDSM 2760), Anchiceratops ornatus (AMNH 5251), Avaceratops lammersi (MOR 692), Torosaurus latus (MOR 1122). Not to scale: the purpose is to show the diversity of postorbital shapes.

Chasmosaurines retained the ancestral condition and went to town with it. While Chasmosaurus lacks impressive postorbital horns, later chasmosaurines are anything but. Most chasmosaurine postorbital horns are tall, and curve gently forward along their length. A few expections exist, including Agujaceratops, whose horns are nearly erect (and if anything, slightly backswept); Avaceratops, whose horns are surprisingly short; and Arrhinoceratops, whose relatively short horns are curved strongly forward. The most impressive postorbital horncores must belong to a giant specimen of Pentaceratops (OMNH 10165), whose horns are measured at 106 cm along their curve. That's 41.7 inches, or just over three feet!

How did chasmosaurine postorbital horns grow? Centrosaurines reabsorbed their horns, but in Triceratops, at least, the postorbitals changed significantly as the animal matured!

Babies have stubs for horns, and these horns grow straight up before curving back in juvenile animals. This posterior curve is retained in subadults, but the horns then sweep forward strongly, more or less at the base, in adult animals. From this point on, the horns actually begin curving anteriorly until they appear subhorizontal in some individuals. Furthermore, as the animal matures, the interior portion of the postorbital horn is reabsorbed! Surprisingly, this complicated reorientation of the postorbitals may be primitive for Ceratopsidae--the same process can apparently be seen in Zuniceratops! Another important point here is that the great chasm in morphology between juvenile and old adult Triceratops provides evidence against Forster's two-species idea. According to Horner & Goodwin's analysis, T. prorsus is merely a subadult T. horridus.

Do the shape or orientation of the postorbital horns have any phylogenetic significance? Sadly, probably not. Because their shape changes so radically with growth (in all ceratopsids), their taxonomic significance is negligible. What appears to be a unique feature, like the strongly angled horns of Arrhinoceratops, is more likely age-related. Even the mere presence of postorbital horns, which at one point separated chasmosaurines from centrosaurines, is no longer a viable phylogenetic feature, since basal centrosaurines also have long postorbital horns, as did the common ancestor of both groups. Even the presence of a postorbital boss vs. horn may not be significant--juvenile centrosaurines, including pachyrhinosaurs, have small horns while the adults have bosses. This puzzle is perhaps best exemplified by "Brachyceratops," a small juvenile centrosaurine from Montana. Its child-like features include small, slightly recurved postorbital horns and an unfused nasal horn. The structure of its squamosal bone gives it away as a centrosaurine, but given what we now know about postorbital growth, its horns hold little value in determining its classification. Generally, the structure or presence of postorbital horns is helpful only after other factors are considered.

What did ceratopsids use their postorbital horns for? The obvious answer is "defense!" And indeed, I doubt Triceratops turned its horns away from an attacker. However, their primary purpose may have been signaling and/or intraspecific combat. The fact that different age classes in Triceratops look different means that individuals could tell how each other animals were at a glance, or how experienced they are. And look at African antelopes--subtle differences in horn morphology reflects differences in taxonomy. Although between you and me, the ceratopsian differentiator was probably parietal morphology (we'll look at that later). Still, within a group of animals, it would have been very useful to identify individuals' ages and ranks at a glance, which differing postorbital horn morphology can do.

Andrew Farke has argued convincingly that ceratopsids with long postorbital horns may have used them for intraspecific combat. He first played figured out how Triceratops was able to lock horns, as it were, and then went out to the fossils to look for signs of pathology that would validate (or tear down) his idea. Turns out Triceratops skulls show numerous injuries to the squamosal bones, just where the point of the postorbitals would be pressed in a fight. So aside from perhaps defense and intraspecific signaling, ceratopsians with long brow horns used their postorbitals for intraspecific combat, as well! Gotta love practical experiments!

So what have we learned today? Well, large postorbital horns are plesiomorphic for Ceratopsidae, but aside from that, the animals within the two groups (Chasmosaurinae and Centrosaurinae) differ wildly in their expression of those postorbital horn morphology. Not only that, but the postorbitals of all ceratopsids change radically throughout life. Postorbital horns would have had many uses, including defense, intraspecific signaling, and most interestingly, intraspecific combat.

Next up: Nasal horns!

P.S. I'm too tired to write the whole big reference list right now. I'll just copy and paste it from my electronic index later. For now, trust that I'm not making anything up.


Andy said...

Awesome post (on a topic near and dear to my heart)! Re: the T. prorsus vs. T. horridus thing, the holotypes for both species are old adults. . .I'm quite skeptical of claims synonymizing them as ontogenetic stages, to say the least. Judging by the NAPC abstract volume, we should have some interesting contributions on Triceratops coming out of MOR in the near future, though!

Ivan said...

Great write-up.

In your post, you make it seem as if Avaceratops is a chasmosaurine though.

Metalraptor said...

You may want to watch out for the idea that T. prorsus and T. horridus are the same species. The idea has been proposed by Horner, who is a notorious lumper, especially when it comes to Maastrictian taxa (partially since he wants to have as little dinosaur diversity as possible in the Maastrictian to go along with his theories). But I agree, the evidence for T. prorsus and T. horridus seems rather strong here. But if both are old adults, this may complicate things.

Anyway, it also does seem more likely that chasmosaurines used their horns for defense or intra-specific combat as much as display. Centrosaurine's horns came in a wide variety of forms, twisting this way and that. Compared to them, chasmosaurines are unmistakably vanilla. They all have pretty much the same two postorbital horns, one nasal horn combo, and they mostly point the same way. This probably indicates that that position was useful for something. Combat seems the most likely reason. Not to mention that Triceratops has a solid frill amongst ceratopsians. Not saying that the frill was for defense, but one would expect a stronger frill on an animal that used its horns for defense, to keep the frill from breaking like a puny centrosaurine frill when it bashed heads.

Not to mention that some research suggests that Triceratops did use its horns for some kind of combat. Thought I am not sure how good this research is, since I read an article once (was it in Carpenter's books I think?) that suggests the disputed "horn holes" were either pathological from some disease, or unique individual marking (sort of like a birthmark).

Speaking of lumping and the Maastrictian, I just thought of some odd things. First, people oftentimes claim that there was something wrong with the Maastrictian biota, since Triceratops was the most common animal. But ceratopsians skulls are notorious for fossilizing well, especially Triceratops of all species. Maybe the Maastrictian biota wasn't that biased, it was just a preservational bias towards ceratopsians. Second of all, I cannot believe that someone has not tried to synonymize all of the chasmosaurines and all of the centrosaurines yet. I know that this would be horribly wrong, but I talked with Dr. Michael Ryan of Paleoblog about this once, and he said that postcranially ceratopsians are pretty much indistinguishable from each other. So I could see someone actually trying to lump them together, even though this would never be accepted in the paleontological community.

Zach said...

Thanks for the comments, guys. Andy, I didn't know they were both old adults! The skull Forster uses in her "two-species" paper (YPM 1922) looks like it confirms nicely to Horner's young adult stage, what with the slightly upturned postorbitals and more obvious epoccipitals.

Are the two species geographically disparate? Could this be a case of sexual dimorphism? Hell, I'm a lumper too. :-)

Metalraptor: As far as I know, nobody's tried synonymizing entire groups--I assume you mean at the generic level. Centrosaurus used to include like three different genera. Otherwise its taxonomy has remained pretty stable. Agujaceratops used to be a species of Chasmosaurus, though.

Metalraptor said...

I mean like pulling a Paul and lumping all members of Centrosaurinae in Centrosaurus, and all members of Chasmosaurinae in Triceratops (or whichever genera came first in these cases).

Andy said...

metalraptor said: Not to mention that some research suggests that Triceratops did use its horns for some kind of combat. Thought I am not sure how good this research is, since I read an article once (was it in Carpenter's books I think?) that suggests the disputed "horn holes" were either pathological from some disease, or unique individual marking (sort of like a birthmark).Both papers involve the same authors (I'm senior author on one, and junior author on the other), talking about different phenomena. The "horn hole" paper (in the "Horns and Beaks" volume) is talking about the rather common occurrence of fenestrae in the squamosal bone of the frill in various ceratopsids. We (Darren Tanke and I) think that most occurrences of squamosal fenestrae are not pathologic, because 1) they are reasonably regular in shape, size, and position (nicely rounded, relatively small, and always in the middle of the squamosal); and 2) the fenestrae almost never show any sign of trauma (healing fractures, osteomyelitis, etc.). It is difficult to imagine a wound (short of a laser beam from those extraterrestrials that were the true cause of the K/T extinction) that would leave such a clean piercing through the frill. Instead, we think these holes are probably just the result of random bone resorption processes - you see similar things going on in thin bones of other animals (e.g., scapulae).

The other paper to which you refer, on "combat" in Triceratops, does not necessarily conflict. In that paper, we (me, Darren, and Ewan Wolff) didn't deal with these extra fenestrae at all, but instead dealt with things we felt more likely to be true trauma (from whatever source) - fractures and periosteal reactive bone (often resulting from "scrapes"). Here, we do suggest that the pattern is consistent with some form of combat behavior (unlike what is seen with the fenestrae).

So in summary, most of the extra holes in the frill probably aren't the result of trauma, but fractures and scrapes probably are.

Re: Zach's question on geographic disparity, etc., this is a point of some discussion right now. I would definitely refer you to the NAPC abstract volume (available as a PDF), to have a look at Scanella and Fowler's work.

Paleo King said...

Lol! For lack of a better word, Horner's theories are a total GAS.

Horner predicted, among other things:

* T.rex was a scavenger at heart (disproven by partially healed T.rex bite marks on Triceratops skulls)

* Nanotyrannus was just a juvenile Albertosaur (disproven when Bakker chiseled through the false plaster preorbital horns)

* Dracorex and Stygimoloch and all the other small boneheads were all juveniles of Pachycephalosaurus (disproved by complete fusion of skull bones in the smaller genera, and by simple morphology - plus such a radical metamorphosis is not seen in anything more advanced than an Ocean Sunfish)

* AND NOW... "Triceratops horridus and T. prorsus are the same...." and he fiercely defends these baseless hypotheses against far more rock-solid ones. When will these inane lumper ego trips end?

The only reason he's such a celebrity is his work on Maiasaura - which he wasn't even interested in to begin with. I think Triceratops actually had THREE species (I count ?T.elatus? because of its more flattened head and lack of upward frill curvature, and large size).

Ironically, Horner's bid for less maastrichtian diversity is an attempt to fit into his rival Bakker's theory of disease wiping out the dinosaurs (lack of diversity is a possible sign of a pandemic) - but even so, three Triceratops species is STILL not a lot of diversity. So Horner is, in my mind, trying FAR too hard.