Thursday, December 01, 2016

Meet the Tyrants, Part II: Intermediate grade basal tyrannosauroids

Yeah, that's a mouthful.  It's been a pet peeve of mine for many years that cladistic methodology has arbitrarily decided that paraphyletic groups are out, since paraphyletic labels are often incredibly useful and convenient.  How many times, if you read dinosaur books, do you hear references to prosauropods, thecodonts, hypsilophodonts, or for that matter, the word dinosaur itself?  All of the time.  Often, sadly, with a pedantic and somewhat self-righteous caveat that you "shouldn't" because it's "imprecise."  Actually, you could construct paraphyletic groups that are just as precise as monophyletic ones—except that of course nobody does because of the arbitrary rule that you shouldn't.

Be that as it may, there is no paraphyletic label to refer to intermediate grade basal tyrannosauroids because nobody had described any or even knew that the group existed except as some kind of ghost lineage, prior to the advent of phylogenetic analysis.  Heck; as recently as Greg Paul's Predatory Dinosaurs of the World he was still calling tyrannosaurs carnosaurs and presuming that they were descended from the allosaurs or something.  Lacking a preexisting label, and unwilling to type out repeatedly "intermediate grade basal tyrannosauroids" I'm going to just grab the last syllable of that big mouthful and call them 'roids here in this series.  What are the 'roids, exactly?  Essentially they are the members of the tyrannosaur family that don't cluster as part of the early proceratosaur radiation (see last post) and also don't qualify as part of the more derived tyrannosaurid clade either (see next two posts).  So they're intermediate grade; in between the early radiation and the later radiation, but they're kind of strung out rather than forming their own "resolved" family.

As I said last time, I'm using the David Hone cladogram from the book I just read, with only a few minor changes.  That seems to have been based on the Brusatte and Carr 2010 study, for the most part.  Hone also refers to the Loewen et al. 2013 cladogram, which is what Wikipedia has, for whatever reason, as its default cladogram. Brusatte and Carr did a new 2016 cladogram; actually two; based on two methodologies, that give very similar results to each other.  I suspect that, although they're pretty new and haven't yet "sunk in" that these will end up being the standard for some time.  I probably should have used a composite of them instead of the earlier Hone one, but for reasons of convenience to me, the Hone one does the job well enough—besides this whole four post blog series was kicked off because I read Hone's tyrannosaur book.  I'll refer, in the specific animal entries, to where it differs from what was more recently published.

Aviatyrannis jurassica. This little guy is often missed from the cladograms (he's not included in Hone's) and even the new Brusatte and Carr study only includes it in the Bayesian cladogram, not the parsimony cladogram.  Known only from a tiny partial hip, not much can be said about this guy except that he's very small; probably about 3 feet long and as little as 12 lbs. or so—although it may well not have been fully grown.  When I was a kid reading every dinosaur book I could get my hands on, Compsognathus was routinely called "the smallest dinosaur" even though he was larger than this specimen.  (Of course, many smaller dinosaurs have been found since then, especially from some of those ashy lacustrine Chinese formations.)  The Bayesian cladogram calls out Aviatyrannis as a basal—in fact, the most basal—'roid outside of the proceratosaur family.  It was originally referred to Stokesosaurus because of its obvious similarities.  Keep in mind that in the new Brusatte and Carr cladogram, Stokesosaurus and Juratyrant fall outside of proceratosaurus, and just up the tree from Aviatyrannis.  Aviatyrannis is from the Alcobaça Formation, which I don't know very well and can't readily find much information on, from the Kimmeridgian (155-150 or so million years ago) from Portugal.  Although I can't find much info on the Alcobaça Formation, the Lourinhã formation from nearby at late Kimmeridgian and Tithonian (150 to 145 or so million years ago) is remarkably similar to the Morrison formation, even having many of the exact same genera (Torvosaurus, Allosaurus, etc.) and when they're not the same, they are extremely closely related and similar (stegosaurs, brachiosaurs, camarasaurs, diplodocids, hypsilophodonts, early iguanodonts, etc.)  Most likely Aviatyrannis occupied a similar faunal assemblage and ecological niche as its close relative Stokesosaurus.

Dilong paradoxus.  Although a 2014 cladistic study (Porfiri et all) found this within Proceratosauridae, every other cladogram has Dilong as one of the earliest 'roids.  That isn't to take away from the fact that it was very similar in many ways to the proceratosaurs; long, three-fingered arms, relatively unspecialized ankles and legs, etc.  Dilong had the typical tyrannosaur pubic boot, though, and lacked the crests of the proceratosaurs.  Most famously, it was feathered; with a kind of filamentous, "woolly" feathery covering.  At only about 5 feet long and probably no more than 35 lbs or so, Dilong was the size of a medium sized dog.  Coming from the 125 million year old Barremian or earliest Aptian stage Yixian Formation of China, it's also among the most recent of the animals we've reviewed so far (excepting Sinotyrannus who came from the layer immediately above it).  The Yixian is famous for its preservation of small biota—not that such small biota was unique or even unusual, but rather that preservation of them is.  It's a big question how much of the Yixian type biota can be applied more generally across wide swaths of the Mesozoic for which no comparable small critters are known, but probably it can be at least in broad strokes if certainly not with regards to the very specific species found.  The environment is lacustrine, so animals like frogs and toads and loads of fish are well known.  There are a lot of small mammals, and the Yixian has gone a long way towards clearing up all kinds of questions about the state of mammals during the Early Cretaceous; many of the specimens are from lineages which are poorly known elsewhere, like gobiconodonts, eutriconodonts, metatherians, etc. although some more traditional mammals, including a multituburculate and at least one eutherian are known.  Lots of lizards, turtles and pterosaurs are also known.  Heck, we even know a bunch of bugs from the Yixian.

Perhaps most famous from the Yixian fauna, however, is the vast assemblage of otherwise completely unknown primitive birds and bird-like maniraptorans—troodontids and deinonychosaurs, often tiny fliers and gliders and other small creatures.  Not only is the fact that so many of these tiny creatures having been found unusual about the Yixian, but the fine preservation with many feather imprints for many creatures is very unusual and makes the formation justly famous.  Larger creatures include some oviraptorosaurs, therizinosaurs, small bipedal ceratopsians, compsognathids, primitive ornithomimosaurs, small iguanodont-grade ornithopods, etc.  Very little is known about the larger animals, which is very atypical, but there is a medium-sized early hadrosaur (about 20-25 feet long) and some poorly known and probably on the smaller side (based on the remains we have so far, at least, which of course may not have been adults) basal titanosaurs.  All in all, the Yixian is a very unusual formation in terms of preservation bias, but it probably was rather more prosaic that we think, it's just that because it has a very different preservation bias than most, so we see a glimpse into a slice of the Mesozoic world that was commonplace, but rarely preserved in enough detail for us to perceive it elsewhere.

Yutyrannus huali. This is the largest of the animals we've surveyed so far.  In the Hone cladogram, it comes next on the 'roid line after Dilong, but Brusatte and Carr made a strong case for it being a proceratosaur, clustered very closely with Sinotyrannus (on the other hand, they stick Stokesosaurus and Juratyrant together about where Yutyrannus is on the other cladograms.)  Although Yutyrannus is large; allosaur-sized, so around 30 feet long and probably weighing as much as a rhino, it really looks more like a scaled up Guanlong rather than a close analog to the more derived tyrannosaurids.  True; it had a shorter neck and larger head, but that comes with the territory when you're a large, probably apex predator for the region.  Most famously, of course, Yutyrannus is known for being feathery in spite of it's large size.  Coming from the same formation as Dilong, it was pointed out that the Yixian was probably colder during the winter than many other formations surveyed, with freezing temperatures and snow—not unlike continental, temperate forest zones of today.  This may at least partly assist in explaining so many specimens found seems to be so "fuzzy."

Known from three nearly complete skeletons; one adult, one subadult and one juvenile, the animal is relatively well known.  As Brusatte and Carr point out, the proceratosaur analogs are multiple (except for the size)—it had narrow teeth, a fairly lightly built head with little in the way of post-orbital bosses and horns, it had a snout crest (more modest than Guanlong's, but not nothing), unspecialized legs, long arms with three-fingered hands, etc.  In short, it's interesting to see that the tyrants were starting fairly early to reach for apex predator status in their faunas, but because it comes from a "primitive" cluster, it's not really very fair to compare them to the very derived and specialized forms of the more "classic" tyrannosaurs.

Yutyrannus "pack" within its natural setting.
Eotyrannus lengi. One of the earliest described and recognized tyrannosauroids, Eotyrannus went a long way towards starting to clear up all kinds of questions about how the tyrants formed, and where.  It was believed prior to this discovery that the tyrants had an Asian origin—this was before the discovery of all of the Asian feathered specimens, and before the recognition of Proceratosaurus and Stokesosaurus as members of the group.  Eotyrannus comes from the exact same time as Dilong and Yutyrannus, but from the opposite side of the Eurasian continent, in England's Wealden Group on the Isle of Wight.  The specimen is known from one fragmentary but intriguing skeleton that is probably a juvenile or subadult about 13 feet long or so—so the full size of the adult animal is unknown.  Some of the very classic Barremian-Aptian European dinosaurs are known from the same formation: Hypsilophon, Iguanodon, Polacanthus, some poorly known diplodocid and rebbachisaurid sauropods, etc.  Predators of the fauna are a little more poorly known, or at least were until more recently, but some spectacular discoveries in the last decade or two have actually opened up various previously unrecognized families, such as Baryonix, Neovenator, and... of course, Eotyrannus.  Small predators probably include, although the material found is difficult to interpret, various maniraptorans, dromaeosaurs, compsognathids, etc. In this regard, the fauna is a bit different than what we see in Asia, and more like what we see in North America at this time—a large carnosaur (Acrocanthosaurus), several sauropods of lingering Jurassic vintage lineages (Sauroposeidon), small, medium and large sized ornithopods, a spinosaurid, etc. Or, for that matter, it's not terribly different from Gondwana-style faunas as found in South America and Africa from the Aptian-Albian either.  But all of this was about to change; sauropods were going to fade in the northern hemisphere, carnosaurs were going to become much more rare if not completely absent, ankylosaurids and ceratopsians and more derived ornithopods like hadrosaurs were all on the rise.  And hunting them, instead of carnosaurs, were going to be tyrannosaurs.

Of course, we're not quite sure how we got there yet.  Sadly, there is a pretty big gap between these last periods of the Early Cretaceous and the last periods of the Late Cretaceous.  We actually know relatively little about the Cenomian, the Turonian, the Coniacian or the Santonian periods; a gap of nearly twenty million years.  There's one more early, somewhat primitive 'roid left to talk about before we hit on this gap, though.

First; one other curious fact.  Porfiri et al did a cladogram in 2014 that recovered Eotyrannus as a megaraptor.  Otherwise, the megaraptors, while phylogenetically a bit bouncy in general, are assumed to be derived neovenators, i.e., late-appearing cousins of Allosaurus.  Based on this, there's been a bit of a swan song floating around in paleontological circles in general that the megaraptors might be a derived radiation of tyrannosauroids of some kind.  I think this is almost certainly a red herring.  The tyrants have been a bit like a Pied Piper of sorts for drawing comparisons with any other therapod—and even occasionally some thecodonts: Sankar Chatterjee once implied that the tyrannosaurs were actually direct descendants of rauisuchians like Postosuchus.  A different carnosaur group, the carcharodontosaurids are often claimed based on one or more similarities to possibly be related to the tyrants, and abelisaurs are as well—but this is the cart coming before the horse, as based on many other synapomorphies, the carcharodontosaurids are a yet different radiation of late-appearing Allosaurus cousins and the abelisaurids are very late appearing descendants of Ceratosaurus.  I had thought, briefly, of cataloging the megaraptorans as part of this catalog because of that claim, even though I don't personally take it very seriously, but I'm now thinking it makes more sense for me to eventually do a catalog of carnosaurs too—my other favorite lineage of big dinosaurian therapods.

Xiongguanlong baimoensis.  This medium-sized carnivore is from China's Aptian-Albian period (125-100 million years ago.)  The exact age isn't clear, but I'm angling towards the earlier end of that big 25 million year range as being more likely; which makes the problem noted just above even worse; instead of a nearly 20 million year gap, it could be closer to 45.  There are still a number of 'roids left to cover, but after Xiongguanlong they are much later in time, and cluster in terms of features much more closely with the derived tyrannosaurids.  Although Xiongguanlong is about the same size and similarly built to Eotyrannus, with a relatively gracile form, longer neck, smaller head, long arms, three-fingered hands, etc.—and like them, in the Brusatte and Carr cladograms, he clusters near the proceratosaurs and represents a similar "grade" of 'roid—he also includes the first appearance of some "tyrannosaurid grade" developments; strengthened skull with some specific struts for muscle attachment, broadening of the base of the skull, ventrally convex maxillae, assymetric serrations on the maxillary teeth, etc.  This is kind of obscured by the fact that it also had a very long snout, quite unlike the tall and deep snout of most derived tyrants, and also lacked the D-shaped spike-like teeth in favor of narrower, more "knife-like" teeth that earlier forms had.  The earlier theory that tyrants evolved in Asia is probably not true; we have proceratosaurs and other primitive forms all over the place, but the specific features of the derived forms might well have, as Xiongguanlong is the first to really have some of them.  However... we don't really know a lot about what followed Xiongguanlong as we don't have any other 'roids until we get to the Campanian period of 80 million years ago that are sufficiently complete that we can tell much about them.

Bagaraatan ostromi.  Brusatte and Carr didn't even include this specimen, although Hone and Loewen do—and probably for good reason.  We really don't even know if this is a 'roid at all.  Thomas Holtz called it one, but with equal authority it's been called a troodontid and a maniraptoran.  It might even be chimerical, which would explain the difficulty in placing it.  If it was a 'roid, it is in the middle of the gap, from the Nemegt Formation of 70 million years ago in Mongolia.  It's known from partial legs, hip and tail, and a bit of jawbone—with features cherry-picked, it appears, from all over Therapoda from birds to even ceratosaurs.  It would have been relatively smallish; probably smaller than Eotyrannus, but other than that, it's probably not worth talking too much about because it's more confusing rather than helpful to consider it.  If it is a 'roid, it doesn't really seem to be indicating much in the way of development towards the derived tyrants, so it might have been a weird off-shoot doing it's own thing.

Raptorex kriegsteini.  This is another problematic specimen.  While undoubtably part of the tyrant lineage, exactly what it is or even when it is from is questionable.  It was initially interpreted as from the Yixian formation (sadly, it wasn't excavated by professionals, so the details have to be guessed at.)  It was later decided that it probably came from the Iren Dabasu of the late Campanian, or a "similar" formation.  It's a very small juvenile hindquarters that cannot be confidently paired with any adult specimens, so many have put out calls to declare it a nomen nudum or undiagnosable specimen.  The latest may seem to be that it is an extremely young specimen of an animal very much like Tarbosaurus.

Alectrosaurus olseni.  Another difficult specimen, Alectrosaurus is known properly from a foot, and maybe some very poorly preserved other bits and pieces from the Iren Dabasu formation of Mongolia from the Campanian.  Other material has been referred to it, including some skull and shoulder elements, from the Bayan Shireh formation (of uncertain age—maybe earlier) although this material may not belong to this genus.  So, what we have really is merely a good, solidly tyrannosaur leg from the Campanian.  Early Campanian according to some (83 million years ago) and Campanian-Maastrichtian (roughly 75 million years ago) according to others.  Most workers leave this out of cladograms entirely, because although it's definitely tyrannosaur, it comes up all over the place, with as many as eight equally valid positions, so its relationships and even its specific features are simply too poorly known for it to be very useful.

Dryptosaurus aquilunguis. Originally named Laelaps, this is famous from a beautiful and dramatic Charles Knight painting of two individuals fighting (Leaping Laelaps from 1897.)  It has since fallen out of the public consciousness a fair bit.  It wasn't recognized as tyrannosaurian until 2005 when the discovery of Appalachiosaurus gave it some context.  Those two are the only known large predators from Appalachia; North America was divided by an epicontinental sea into two separate "island continent" landmasses in the Late Cretaceous; eastern Appalachia and western Laramidia.  Sadly, Appalachia is not well known in terms of its fauna.  Other than some hadrosaurs and nodosaurs it's not 100% clear what else lived on Appalachia, but it's curious that nodosaurs (which were otherwise replaced by ankylosaurs on Laramidia) and these two interesting late-appearing 'roids (instead of derived tyrannosaurs) are here, as well as a very different assemblage of hadrosaurs than are found in Laramidia.  There are also some scanty remains that have been referred to ornithomimosaurs and dromaeosaurs, and there even appears to be one very atypical and specialized ceratopsian.  Overall, Appalachia was significantly different than Laramidia, and has a rather unique "island continent" phenomena to it in terms of fauna that survived here that did not in other, more connected continents.

Leaping Laelaps
Dryptosaurus has some features in common with more basal 'roids; a more gracile build generally, smaller, less strengthened head, longer arms, three fingers with long claws (although some proposals suggest that only two were functional in any way), etc.  Curiously, in one of the two phylogenies in the recent Brusatte and Carr paper, Dryptosaurus comes up nestled within Tyrannosaurinae as a close relative of Alioramus.  The authors don't actually believe this result, however—they account for the fragmentary nature of the skeleton thus giving fewer points of comparison as well as the long ghost lineage of the Appalachian 'roids as obscuring exactly what was going on with the radiation in the east, most likely.  It was probably about 25 feet long and weighed a ton and a half.  Although somewhat smallish for an apex predator of the Maastrichtian, as far as we know, it was the largest guy on the block in its locality.

Appalachiosaurus montgomeriensis.  Found in the Turnipseed Dinosaur Site (what a great name!) from the Demopolis Chalk Formation in Alabama, this is the best known eastern therapod, from about 40% of a skeleton.  Although the skeleton indicates an animal about the same size as Dryptosaurus, i.e. a little over 20 feet long, it also is not fully grown and seems to indicate a subadult that was maybe 2/3 of its adult size.  It's from smack dab in the middle of the Campanian, about 77-79 million years ago.  The Demopolis Chalk formation is actually a marine environment and was deposited under a good 100 ft. or so of water.  Many of the dinosaur fossils from Appalachia seem to have been bodies that tumbled down from the foothills or lowland plains in river floods and were washed out to sea, which makes putting them in their proper context a little bit difficult.  A decent-sized hadrosaur (Lophorhothon) is known, as well as some nodosaurian, dromaeosaurian and ornithomimid remains that were probably contemporaneous.

Appalachiosaurus was initially believed to be a very eastward trending Albertosaurus, but subsequent investigation into the remains uncovered its true nature as a derived, yet non-tyrannosauridan tyrannosauroid; or a late-breaking and advanced 'roid.  As I said above, it also provided enough information to compare with Dryptosaurus which led to the belief that Appalachia may have had "island continent" faunal patterns, with "relics" of groups that were otherwise replaced across the Laramidian-Asian intermittently connected landmass.  It also raises questions as to whether or not any connections of note happened between Laramidia and Appalachia—it kind of appears not, or if there were, not very much participated in faunal exchange.

Although it appears earlier, it appears also to be more derived in its own Appalachian 'roid evolutionary path than Dryptosaurus, with a fairly longirostrine skull that has some features in common (against all expectations) with Alioramus from Asia.  The forelimbs are not confidently known, so it's unclear if they were long or reduced, or if they had two or three fingers.  A lot more work, if it can be pulled off, to uncover the secrets of Appalachian dinosaurs would be very welcome.  We're just starting to realize how potentially interesting the Appalachian island continent may have been in the Campanian and Maastrichtian.

Speaking of Alioramus, in the Hone cladogram, it's located next, just outside of Tyrannisaurida and closely related to Appalachiosaurus.  However, that wasn't always the case; in fact, Alioramus was often seen as very tyrannosaurine in nature specifically, but showing juvenile characteristics, such as a long snout, light, smaller build, long legs, etc.  Many even proposed that it was merely a juvenile Tarbosaurus.  With the discovery of a second Alioramus species and the similar (albeit larger) Qianzhousaurus, it was recognized that it was indeed an atypical radiation of early derived tyrants that were lighter, probably faster, and longer-nosed than the deep-snouted classic tyrants, which means that it is not closely related to Appalachiosaurus after all.  The Brusatte and Carr phylogenies show this clearly, and it is one major change that I've adopted to the Hone cladogram that I've otherwise using for purposes of this series of blog posts, in part because it makes it easier for me to lump the alioramines with the albertosaurines for my next post, and then finish up in a fourth post with the classic tyrannosaurines.

The Brusatte and Carr cladogram also removes Bihastieversor from within Tyrannosaurinae and puts it just outside Tyrannosaurida.  The logic here is that the Loewen (and even their own earlier work, which is what Hone is repeating) gave too much diagnostic importance to the "tallness" of the skull.  Once that's corrected for, it makes sense based on the rest of the analysis to remove deep-snouted Bihastieversor from deep within the tyrannosaurines based on the coincidence that it has a tall head, and put the alioramines within the tyrannosaurines in spite of their shorter, longer snouts.  That said, this post is long enough, so I'm going to do Bihastieversor along with the rest of the guys it more closely superficially resembles, the way Hone has written it up in his book.

There are also a few other specimens that are sometimes referred to as belonging to the 'roids, but they are generally known from too fragmentary and non-diagnostic material to really be useful, and most cladograms don't include them because they merely introduce noise rather than helpful info.  These specimens include Labocania from Campanian Baja California, which may be an abelisaurid, or even some other type of therapod altogether, Timimus from Australia's famous Dinosaur Cove locality, which was clearly some kind of coelurosaur, but may not have been any type of tyrant (see swan song discussion under Eotyrannus above) and Timerlengia from the Turonian of Uzbekistan.  The latter probably is a tyrant of some kind, presumably a slightly later fairly close relative of Xiongguanlong, which should be exactly the kind of "missing link" both in phylogeny and in time that we'd love to find.  Sadly, the remains are too scrappy to tell us very much; merely two braincases, and some other bones here and there that have been somewhat dubiously referred to the genus.

No comments: