Tuesday, January 30, 2018

Nearly Birds

Adorable Serikornis sungei by Emily Willoughby
Something I really enjoy about paleontology is how quickly things can change. For example, when I was growing up, Dromaeosauridae was confined to half a dozen genera from two continents. If you wanted a complete list, you could check out Raptors: The Nastiest Dinosaurs from your local library. Now, though, Dromaeosauridae is more like Dromaeosauriformes because there are something like five distinct groups now: Halszkaraptorinae, Unenlaginae, Microraptorinae, Dromaeosaurinae and Velociraptorinae (those last two are usually stuck together in a monophyletic Eudromaeosauria). It used to be that dromaeosaurs came in two flavors: large and small. Now you’ve got swan-necked, duck-billed dromaeosaurs; piscivorous, leggy dromaeosaurs; tiny, potentially volant dromaeosaurs; and larger “classic” predatory dromaeosaurs.

All this has happened in the last twenty years. Heck, nobody knew about Halszkaraptor until a few weeks ago.

The holotype of Anchiornis huxleyi
For several years now, paleontologists in China have been digging up theropods that seem to toe the line between “bird” and “non-avian dinosaur.” And many of them are, critically, older than Archaeopteryx. The first of these to be described was Anchiornis huxleyi, named in 2009 by Xuet al. The first described specimen (IVPP V14378) lacked a head and most of the right arm but was otherwise quite complete. Feathers were present but weren’t all that special: “Specifically, extremely faint carbonized feather impressions are preserved dorsal to the presacral vertebral column…but their detailed structure is not clear.” The authors placed Anchiornis right below Archaeopteryx, phylogenetically.

I'm not sure why I'm including specimen numbers in this essay, but I kinda like it.

A much better specimen of Anchiornis huxleyi
Well, it turns out there are many more specimens of Anchiornis out there, both in museums and in private hands, and later that same year, a much more complete specimen was described in Nature by Hu et al. This new specimen (LPM-B00169) provides a much more dramatic picture than the holotype did. Anchiornis is a crow-sized dinobird with long feathers not just on its tail, hands and arms but also on its legs and feet. Hindwings were old news thanks to Microraptor, but feathered feet and toes were new. The authors ran Anchiornis through a phylogenetic analysis (like you do) and found Anchiornis to be the basalmost troodontid (by almost 40 mya), but also that it shares features in common with troodontids, basal birds, and basal dromaeosaurs, indicating that Anchiornis seems to be right at the divergence point for birds and deinonychosaurs.

Also, the Tiaojishan Formation where Anchiornis is found turns out to be older than Archaeopteryx by up to 10 mya.

Xiaotingia zhengi
Next up to the near-bird plate is Xiaotingia zhengi, described in 2011 by Xu et al. It is also a small animal (STM 27-2) from Tiaojishan with faint feather impressions around the whole skeleton, but like Anchiornis, has elongated feathers on its leg and small feathers on the toes. Unlike Hu et al. (2009), who focused on comparisons between Anchiornis and troodontids, Xu et al. (2011) make many comparisons between Xiaotingia and Archaeopteryx, with fewer comparisons to Anchiornis

This paper caused a bit of a stir in the paleo community because the authors’ phylogenetic analysis actually removed Archaeopteryx from Avialae. Instead, Archaeopteryx is now in a group with Anchiornis and Xiaotingia, which is a sister group of Deinonychosauria. This large group, then, is the sister of Avialae. Xu et al. (2011) comment that basal birds like Sapeornis, Jeholornis, and Epidexipteryx (which they seem to think is a basal bird) are “more similar in general morphology to the oviraptorosaurs than to the archaeopterygids and basal deinonychosaurs.”

While the removal of Archaeopteryx from Avialae has not withstood subsequent testing, the idea that Anchiornis and Xiaotingia form a monophyletic group will come up again.

Eosinopteryx brevipenna
2013 saw the publication of two more near-birds, both, of course, from Tiaojishan: Eosinopteryx brevipenna (YFGP-T5197) is the first, described by Godefroit et al. (2013a) in Nature. Like its cousins, it has elongate feathers on its hands, arms, and legs but, notably, feathers seem to be absent on the metatarsus and toes. Tail feathers are short, which is not the case in some of the other near-bird taxa we’ll discuss. While Eosinopteryx is smaller than Anchiornis or Xiaotingia, it appears to have died as either a subadult or adult, so Godefroit et al. (2013a) do not believe that the lack of a complete hindwing is ontogenetic. Their phylogeny, however, puts Anchiornis, Xiaotingia, and Eosinopteryx back at the base of Troodontidae while leaving Archaeopteryx (and Wellnhoferia—a topic for another day) at the base of the Deinonychosauria. This major group is, again, a sister to Avialae.

Godefroit et al. (2013a) posit that Eosinopteryx was more terrestrial than its cousins, both because of the featherless metatarsus and feet and its smaller wings. Interestingly, they note that Anchiornis and Eosinopteryx are very similar osteologically but clearly differ in terms of plumage. We may come back to this point later.

Aurornis xui
And then, not too much later, Godefroit et al. (2013b) described Aurornis xui. While feathers are not very well preserved on this specimen (YFGP-T5198), the skeleton is certainly in better shape than Xiaotingia and it looks a whole lot like Eosinopteryx. Can you guess where it's from? That's right, kids, Tiaojishan. There’s not too much to say about its anatomy that hasn’t already been said about its brethren, but the main takeaway (for me) is that their new phylogeny shuffles the all these basal paravians around again. None of the previously-described near-birds are troodontids anymore. Instead, Eosinopteryx is the ougroup of Paraves and everyone else forms a stepwise progression within Avialae in this order: Aurornis, Anchiornis, Archaeopteryx, Xiaotingia, and then all other birds.

But it’s clear that nobody can agree who is actually related to whom because all of these near-birds are so much alike. Frustrating, right?

A way, WAY better specimen of Anchiornis huxleyi
During all this time, Anchiornis was quietly becoming super well-known in terms of osteology, soft-tissue, and plumage. In 2017, Pei et al. described four beautiful new specimens of Anchiornis in a gorgeous monograph that you should all read if only for the jaw-dropping photos. Those authors confirm that Anchiornis is a lot like the other Tiaojishan near-birds (surprise!), and that it shares many features with Archaeopteryx that are not found in deinonychosaurs. As for its troodontids characteristics, Pei et al. find those to simply be more commonly distributed among basal paravians and are not unique to troodontids, so there's no good reason to call them troodontids. The authors suggest that Anchiornis is a basal avialan and, furthermore, that the Tiaojishan near-birds “likely form a monophyletic group based on their anatomical similarities.” Interestingly, Pei et al. (2017) regard Aurornis as a junior synonym of Anchiornis but Andrea Cau has debated that.

Serikornis sungei
Yet another near-bird was described in the middle of 2017 by Lefevre et al. Serikornis sungei is yet another small Tiaojishan taxon (PMOL AB00200). While also quite small, Serikornis differs considerably from its cousins in the fine details of its skeleton but also its proportions and plumage. It would seem that all of these near-birds were covered with a surprising variety of feather types, and that their lack of unambiguously scansorial or arboreal adaptations suggests they were largely terrestrial. In their phylogeny, Lefevre et al. (2017) find that Aurornis, Eosinopteryx, and Serikornis form a monophyletic outgroup of traditional Paraves, and Anchiornis is the basalmost paravian. What about Xiaotingia? Their results oddly place that taxon in the Scansoriopterygidae but “its position is highly unstable and further investigations are required to fix its position.”

Then, at the end of 2017, Foth & Rauhut took a closer look at the Haarlem Archaeopteryx fossil and realized that it’s probably not an urvogel, but an anchiornithid instead, which they named Ostromia crassipes. Furthermore, they actually give a name to this group of near-birds for the first time: Anchiornithidae. Their phylogenetic analysis does, in fact, place Anchiornis within the Anchiornithidae (contra Lefevre et al. 2017) which includes all the other near-birds I’ve been discussing plus Ostromia, of course. Interestingly, they find that Epidexipteryx is not just an avialan, but is one step higher than Archaeopteryx

(you know, what we really need are more, and better preserved, scansoriopterygian fossils)

It’s also notable that Ostromia is the first anchiornithid to be found outside of China, although another one may be described before too long based on Scott Hartman’s SVP talk…

Caihong juji
And now, just last week, another small Tiaojishan near-bird came out of the pipeline—Caihong juji is the most divergent of the group (Hu et al. 2018) (PMoL-B00175). Its skull is elongate, looking more like Velociraptor than Anchiornis. It may or may not have lacrimal horns (they might be a distortion artifact). Its arms are proportionately shorter than other Tiaojishan theropods but, weirdly, the ulna and radius are proportionately longer. In fact, it is longer than the humerus, “a feature so far known only in flighted avialan taxa among theropods.”

Beautiful feather impressions surround the whole body. Caihong’s wing feathers are much longer than Anchiornis and it appears to have an alula. Its leg feathers are also quite long, and like Anchiornis it has feathers on its metatarsus and toes. The authors note that:

Eosinopteryx brevipenna has been suggested to have reduced tail and hindlimb plumages, but specimens of Anchiornis huxleyi display variable plumages in terms of not only feather distribution, but also feather size and shape. Some closely related Tiaojishan theropods not differentiated by osteological features may need reassessment of their taxonomic status.

I'm not sure any of these Tiaojishan anchiornithids have been described as identical apart from plumage differences. The tail feathers of Caihong are particularly long and, like the basal troodontids Jianianhualong, appear to be asymmetrical; this is especially strange given that the wing and leg feathers are not. Like Serikornis, Caihong features a wide variety of feather types on its body. Its preserved melanosomes suggest that Caihong was mostly black, but featured hummingbird-like iridescence on its head, chest, and possibly the base of the tail. The only other iridescent dinobird I know of is Microraptor (Lu et al. 2012), but its melanosomes were not directly compared to hummingbirds. The implication of Hu et al. (2018) is that Caihong was colorful--moreso than its contemporaries, anyway.

The authors’ phylogenetic study resembles that of Xu et al. (2011) in that Archaeopteryx is removed from Avialae and instead forms an outgroup to Anchiornithidae + Deinonychosauria. However, all of the anchiornithids are found to be monophyletic.

Rebecca Gelernter's Anchiornis
And that's where things currently stand. Another interesting thing about Tiaojishan is that it seems to be dominated by pterosaurs and small mammals. Aside from the animals discussed here, the only other dinosaurs present are all three scansoriopterygian taxa and Tianyulong, the fluffy heterodontosaur. Where are all the other dinosaurs? Tiaojishan is older than the famous Yixian Formation, which is where every other famous feathered dinosaur is from. That there are no deinonychosaurs or avialans at Tiaojishan seems to support the idea that anchiornithids predate the divergence of both.

Another oddity is that there apparently hundreds of specimens of Anchiornis available for study but every other Tiaojishan anchiornithid is known from a single specimen. What made Anchiornis so successful compared to its neighbors? How capable were these animals from an aerodynamic angle? Did they make frequent trips up tree trunks and then glide back down or between trees? The extreme feathering on the legs and feet of several anchiornithids doesn't seem terribly useful in a terrestrial context. Rebecca Gelernter restores Anchiornis in a hypothetical branch-climbing pose, but would it have been able to do this in life? Would any anchiornithid? Besides, with an apparently lack of carnivores to run from, would anchiornithids feel any pressure to get off the ground at all? How did they interact with the numerous pterosaurs and small mammals who were their neighbors?

Whatever the answers may be, I suspect that anchiornithids have much to tell us about the origin of birds, modern plumage, and (maybe) flight. And with such beautiful fossils coming out of Tiaojishan, the answers may be closer than we think.


  1. "The extreme feathering on the legs and feet of several anchiornithids doesn't seem terribly useful in a terrestrial context."

    Not extremely useful in an arboreal context either, since said feathers would snag on branches (a point made by Dececchi & Larsson, 2011).

    Long hindlimb feathers were undoubtedly not advantageous for running. But these 'nearly-birds' might not had much need for cursoriality, even if (as seems likely) they were terrestrial. It is difficult for very small, long-limbed, cursorial animals to be too fast, because obstacles on the ground become relatively larger the smaller you get.

    "Rebecca Gelernter restores Anchiornis in a hypothetical branch-climbing pose, but would it have been able to do this in life?"

    No. ;-) (Well, I don't think so.) I love her artwork, but the arboreal restoration looks very 'forced' to me.

  2. Thanks, Tim! I really enjoyed researching it. 😊