Using the same 231 characters from
the large reptile tree (LRT, 1085 taxa) the subset of extant birds and their allies also came out fully resolved (Fig. 1).
Prum et al. 2015 used DNA
to determine the phylogeny of Neoaves (nearly all living bird species). They reported this “remains the greatest unresolved challenge in dinosaur systematics”, but that was before the recent alignment of Ornithischia and Theropoda by Baron et al. 2017.
I have to admit
as usual, before I started adding more extant birds to the LRT, I knew nothing about them. Their generic names were new to me. You might remember the LRT started with just a chicken (Gallus) and an ostrich (Struthio). Now there are 42 birds with 143 outgroup taxa.
Birds are tough.
Often they fuse skull bones. That may be why other workers find protrusions and bumps to base their traits on. Some of the best data for many taxa come from decades old drawings and photos from skull sellers. I made many mistakes along the way, now minimized. The cladogram was my mentor here, telling me with autapomorphies where to look for mistakes.
Matching all prior workers,
tinamous and ratites were recovered as basalmost taxa. In the Prum et al. DNA study, chickens, crakes, screamers and ducks branch off first. In the LRT, which includes extinct taxa, the predators and toothed birds split off first. Distinct from the Prum et al. study, in the LRT long-legged walking birds are basal to many clades. Even the basalmost toothed bird, Yanornis, from the Early Cretaceous, is a long-legged walking bird, also capable of flying. And yes, this puts the origin of the clade of extant birds back to just after the Jurassic. Jurapteryx, from the Late Jurassic, is not far off.
Herons come next,
followed by all other birds with the corn crake (Crex) the hammerkop (Scopus) and the limp kin (Aramus) at the base. Adding taxa allows me to amend an earlier nesting of the elephant bird (Aepyornis) with ducks. Now Crex and Aepyornis nest together.
In the Prum et al. study
swifts + hummingbirds split off after chickens + ducks.
in the LRT swifts (Eocypselus) and hummingbirds (Archilochus) nest between terns (Thalasseus) and kingfishers (Megaceryle). Nearby, high-energy dippers (Cinclus) nest with other wing swimming birds: murres (Uria) and penguins (Aptenodytes). Cinclus is traditionally considered a passerine, but the sparrow, Passer, does not nest with it in the LRT. Passer nests between chickens and parrots (Ara), all seed eaters.
In the Prum et al. study
seed-eating passerines arise from carnivorous falcons and seriema (Cariama). That does not seem right on the face of it. In the LRT passerines arise from omnivores, like Chauna.
juvenile traits found in adult specimens, evidently produced all of our short-legged birds and produced smaller adult birds, found at derived nodes. Juveniles of flamingos and other long-legged taxa have short legs. Of course, some small birds also had large and giant descendants, all at derived nodes.
As in many studies that conflict with the LRT
the lack of appropriate fossil outgroup taxa seems to set their cladograms in other directions. That can happen. DNA studies can never solve this problem.
Apologies for earlier mistakes due to
naive misidentifications and taxon exclusion. Those will be repaired.
Baron MG, Norman DB, Barrett PM 2017. A new hypothesis of dinosaur relationships and early dinosaur evolution. Nature 543:501–506.
Prum RO et al. (6 coauthors) 2015. A comprehensive phylogeny of birds (Aves) using targeted next-generation DNA sequencing.