With recent advances in genetic sequencing and analysis, we now have a pretty good idea of how most modern vertebrate animals are related to each other. One of the biggest remaining mysteries in vertebrate evolution (and a major theme of my own research), however, is the relationships among the major groups of living birds.
There are some things that we all agree on about the bird family tree (which in some cases were already recognized before the rise of genetic studies), a big one being that modern birds can be divided into two major branches: Palaeognathae (ostriches, emus, and their close relatives) and Neognathae (all other living birds). Neognathae is in turn divided into Galloanserae (chickens, ducks, and their close relatives) and Neoaves (all remaining birds, which constitute 95% of living bird diversity).
Despite birds being one of the most intensely studied animal groups, however, essentially none of the large-scale genetic analyses that have been done on them so far have agreed with each other regarding how the major groups within Neoaves are related.
A new study by Stiller et al. (2024) might represent a big step forward in solving this mystery. Their results suggest that Neoaves can be divided into four major groups.
- Mirandornithes: Flamingos and grebes. Stiller et al. found that all other members of Neoaves are probably more closely related to each other than to this group.
- Columbaves: Consisting of two major subgroups, Otidimorphae (cuckoos, bustards, and turacos) and Columbimorphae (pigeons, sandgrouse, and mesites). Notably, Columbimorphae has been found by some earlier studies to be more closely related to Mirandornithes, but a second paper that was published on the same day by some of the same authors as Stiller et al. (2024) reported evidence that this previous result was probably caused by misleading similarities between the genetic sequences of Columbimorphae and Mirandornithes.
- Elementaves: Consisting of Gruiformes (cranes and their close relatives), Charadriiformes (shorebirds), Strisores (hummingbirds, swifts, nightjars, and their close relatives), Phaethoquornithes (many waterbirds, including penguins, albatrosses, and herons), and the engimatic hoatzin. The exact relationships among these groups are still somewhat unclear; for example, Stiller et al. found the hoatzin to be most closely related to gruiforms and shorebirds (as had been suggested by an earlier study), but support for this result was not high. The hoatzin remains the single most difficult bird species to place in the bird family tree. The name Elementaves was newly coined by Stiller et al., referring to the fact that this group includes species specialized for life in the water, on the ground, and in the air (corresponding to the classical elements of water, earth, and air), as well as birds named after the sun ("fire"), such as the tropicbird genus Phaethon (Ancient Greek for "sun") and the sunbittern. This means that there is now a scientific basis for parodying Avatar: The Last Airbender using birds.
- Telluraves: A big group consisting primarily of tree-dwelling birds, including songbirds, parrots, woodpeckers, kingfishers, and the various groups of birds of prey. An interesting result found by Stiller et al. is that owls are likely closely related to accipitrimorphs (hawks, eagles, vultures, etc.), which not all previous genetic studies had supported.
Stiller et al. (2024) provide further evidence for some bird relationships found by earlier analyses, but their results still doesn't exactly match those of any single previous study, so what makes this different from all those attempts that came before it? One is the amount of data. The genetic dataset analyzed by Stiller et al. was many times larger (both in terms of sequence length and the types of genes examined) than any study of this sort that had previously been done on birds. They also included over 360 bird species, which is more than what most previous studies had. Furthermore, they ran numerous tests to determine how the amount of data, number of species, and types of genes analyzed affected their findings, and in doing so were able to show that most of their results were relatively robust, or at least better supported than alternative hypotheses.
Another point of contention regarding the evolution of Neoaves is when the group originated. Were there already many neoavian lineages around during the Late Cretaceous, or did they mostly diversify following the mass extinction event that ended it? In the 2000s and early 2010s, studies trying to estimate the ages of bird groups based on rates of genetic evolution tended to find an older origin for Neoaves, but the majority of newer studies favor a younger origin, with most or all modern neoavian groups appearing after the Cretaceous (though one paper from earlier this year by a different team of authors advocated for older ages). Informed by recent studies on fossil birds, the results of Stiller et al. add further support for a more recent, mainly post-Cretaceous diversification of Neoaves (which I happen to think is more plausible than deep Cretaceous origins).
This almost certainly won't be the last word on these controversies by any means. However, at the moment I'm willing to tentatively consider Stiller et al. (2024) the closest we've gotten to approximating the true family tree of birds, and that is not a declaration I'd make lightly.