Saturday, January 25, 2014

Oh, Hi, Bohaiornithids!

It's not often that we are introduced to a large new clade of stem-birds*, but a new paper by Wang et al. finds support for just such a thing among the enantiornithes. Named Bohaiornithidae, the family unites a few previously-known similar-looking opposite birds with two brand new species.

Phylogeny of Bohaiornithidae, modified after Wang et al. 2014.

The phylogenetic analysis of Wang et al. 2014 found the following to be bohaiornithids: Bohaiornis, Parabohaiornis, Longusunguis, Shenqiornis, Sulcavis, and Zhouornis. All are very similar in overall anatomy, having short, deep snouts, long, wicked-looking talons, and short, conical, slightly curved teeth. Interestingly, Eoenantiornis was also found to be closer to the bohaiornithids than to any other enantiornithes. A more cynical person might speculate it was left out of the group because the names Eoenantiornithidae and Eoenantiornithiformes would both have had priority over the newly-coined name Bohaiornithidae, but that wouldn't be entirely fair, as the eoenantiorn + bohaiornithid clade seems to be rather weakly supported. I provided definitions of both older names in my book, A Field Guide to Mesozoic Birds and other Winged Dinosaurs: Eoenantiornithiformes as all specimens closer to Eoenantiornis than to Cathayornis, Iberomesornis, or Enantiornis, and Eoenantiornithidae as all specimens closer to Eoenantiornis than to Longipteryx, Cathayornis, or Enantiornis. These definitions were chosen based on the phlyogeny of Cau & Arduini 2008, which found Eoenantiornis to clade with the longipterygids. In the bohaironthid phylogeny of Wang et al. 2004, Eoenantiornithidae would be synonymous with Eoenantiornithiformes, which in turn would be the clade containing Eoenantiornis as Bohaiornithidae.

The type species of the bohaiornithids, Bohaiornis guoi, has been known since 2011, based on a specimen reported to come from the Yixian Formation (though later revised to the later Jiufotang formation). Known from a subadult specimen, Bohaiornis was notable for it's broad snout, giving it a superficially dromaeosaurid-like profile, and pair of long ribbon-like tail feathers. Bohaiornis was very similar to the Yixian Eoenantiornis, and I suggested in my Field Guide to Mesozoic Birds that the two may turn out to be synonyms, with Eoenantiornis representing a juvenile growth stage, but this seems less likely now given the newfound diversity of Bohaiornis-like species and the reassignment of Bohaiornis to the Jiufotang.

The second specimen of B. guoi, borrowed from Li et al. 2014. 
A second specimen of Bohaiornis was reported earlier this month by Li et al., in a paper that suggested it might represent a raptorial enantiornithe.** This possibility was based mainly on the presence of gastroliths in the new specimen, the first and so far only known opposite bird to preserve these stomach stones. The gastroliths in Bohaiornis were relatively large in size and few in number, with rough texture rather than polished. This suggests they were not being used in a gastric mill or gizzard, which herbivorous birds and stem-birds use to help grind and process food after it's been eaten.

So why were bohaiorns eating pebbles? If we set aside the possibility the authors mention that these were simply accidentally swallowed during foraging, it's possible that they were used in the kind of pellet regurgitation and stomach purging behavior seen in modern raptors. As anybody who dissected these in elementary school knows, carnivorous birds regurgitate pellets of un-digestable bone and fur or feathers. In some birds, this process also helps to clean out and scour the digestive tract, and some birds swallow roughage specifically to clean out their insides. It's possible that this is why the bohaiorn specimen swallowed those stones. It should also be noted though, that as John Conway and Darren Naish have pointed out a few times on the Tetrapodcats, what a fossil preserves as stomach contents might not necessarily reflect usual diet. We need to remember we're looking at a dead animal, and it's always possible that it was ingesting weird things due to illness or even that something unusual that it ate actually contributed to its death. Maybe the Bohaiornis accidentally got a mouthful of rocks and then keeled over.

Establishing the diet of enantiornithes has been notoriously tricky, as stomach contents of any kind are rare. This alone may suggest they were mainly eating things that don't fossilize easily, like soft parts of plants, or insects and other invertebrates (both of which would fit well with the apparently highly arboreal lifestyle of most Early Cretaceous species). Tooth and claw anatomy should help, but in cases like the bohaiornithids, it simply confuses matters. Bohaiornithids generally have conical teeth that lack serrations, with pointy, slightly curved tips. These robust teeth have been interpreted as an adaptation to eating hard food, like hard-shelled arthropods.

The claw curvature of these enants is interesting. The talons are large and robust, often with an especially robust second toe, somewhat like that seen in dromaeosaurids, but with a standard-sized claw and no evidence of retractability. The keratin sheaths of the claws, where preserved, are very long and gently curved. Wang et al. 2014 compared the claw and leg proportions of bohaiornithids to modern birds, and found that among raptors, only Osprey have similar proportions, though they have more strongly-cured claws than bohaiornithids. So, it's possible that bohaiornithids were semi-specialized fish-eaters, though it would be nice to find some stomach contents to support this.

It should be noted that raptorial enantiornithes have been suggested before. The avisaurids, specifically the large Avisaurus and Soroavisaurus, have been proposed to be raptorial based on their strongly curved talons and orientation of the toes, which are similar to dromaeosaurids and appear well-suited for pinning prey (the ripper hypothesis). These avisaurids were much larger than even the largest bohaironrithid specimens, which were about the size of modern kestrels.

Whatever they were doing, bohaiornithids represent a group of very similar birds (with the exceptions of the Yixian-age Shenqiornis, and of Zhongornis, the provenance of which is unknown) all living together in the same Jiufotang ecosystem, about 120 million years ago. Could it be that all of these similar species are in fact variations or growth stages of one or two species? If they were indeed raptors, having so many species of about the same size class living in the same environment would be unusual, though it's also possible they were all exploiting slightly different niches. They appear to be unspecialized enough at whatever they were doing that some could have been piscivores, some preying on small lizards or insects, etc. Still, I can't help but be reminded of the similarly over-split omnivoropterygids of the same formation...

Whatever, I'm not going to dwell on the minor things, this has been a great month for fossil birds. And I haven't even mentioned the paper arguing that Zhongornis, restored as a baby confuciusornithid in my book, may actually be an oviraptorosaurian scansoriopterygid! It really does look like a confuciusornithid though. Do they have to be mutually exclusive? Could confuciusornithids be advanced basal oviraptorosaurs? Could oviraptorosaurs actually be avialans after all, which would neatly fill in that glaring gap between long-tailed and short-tailed forms? Ok, now I'm going a little too crazy. 'Till next time!

* ...and it's not often I write for DinoGoss lately! Don't let anybody tell you a baby will not become your primary hobby.
** The literature seems to have universally settled on "enantiornithine" for a member of the clade Enantiornithes. This bugs me as a taxonomy nerd, because "enantiornithine" should really refer to a member of the subfamily Enantiornithinae. I'm not sure what should be used instead, though I'm guessing it should simply be "enantiornithe". A monkey is a member of the clade Primates, and we call it a "primate", not a "primatine". Anybody who is a Greek word nerd, feel free to chime in!

  • Cau, A., & Arduini, P. (2008). Enantiophoenix electrophyla gen. et sp. nov. (Aves, Enantiomithes) from the Upper Cretaceous (Cenomanian) of Lebanon and its phylogenetic relationships. Atti della Società italiana di scienze naturali e del museo civico di storia naturale di Milano, 149(2), 293-324.
  • Li, Z., Zhou, Z., Wang, M., & Clarke, J.A. (2014). A New Specimen of Large-Bodied Basal Enantiornithine Bohaiornis from the Early Cretaceous of China and the Inference of Feeding Ecology in Mesozoic Birds. Journal of Paleontology, 88(1), 99-108.
  • Wang M., Zhou Z.-H., O’Connor, J.K., & Zelenkov, N.V. (2014). A new diverse enantiornithine family (Bohaiornithidae fam. nov.) from the Lower Cretaceous of China with information from two new species. Vertebrata PalAsiatica, 52(1): 31-76. (pdf link)


  1. When I submitted my 2008 paper on Enantiophoenix, the first version used "enantiornithean" instead of "enantiornithine", but the reviewers asked for the latter term... :-S
    Nice to see that my suggestion of a "raptor-like" clade among enantiornithines has further support:

  2. O'Connor and Sullivan reach a little bit to affirm their hypothesis. I am looking into it, but it should be noted that scansoriopterygids are probably NOT oviraptorosaurs.

  3. It should be "enantiornithian," just as Aves becomes "avian." But yeah, most other avian workers are too lazy to bother having to reroute a couple of neurons in their brain away from "enantiornithine." One of these days, someone smart will force the issue by naming a formal clade "Enantiornithinae."

  4. I'd be extremely skeptical of results from O'Connor matrices, including this one. They are extremely under-coded (e.g. unknown dentary teeth presence in Shanweiniao). This leads to largely poorly supported nodes in Enantiornithes, as seen by comparing this topology with the O'Connor et al. (2013) Sulcavis tree, or the O'Connor and Zhou (in press) Chaoyangia tree.

    1. The Chaoyangia paper is now published: