A Phylogenomic Tree of Passerine Birds

The OpenWings project has published a new evolutionary tree of passerine birds, a group including nearly 6000 species of sparrows, flycatchers, ravens, chickadees and their relatives. Image courtesy of Daniel Field, University of Cambridge (Flickr)

Passerine birds are birds in the order Passeriformes. The order’s name means roughly “with the form of a sparrow,” passer being the Latin word for that bird. The order includes not only sparrows, but a huge variety of species from tiny kinglets to massive ravens, from the plainest flycatchers to the most ornate birds of paradise. One of the facts about passerines that has intrigued scientists for decades is that this one order (of only about 30 recognized orders of birds) comprises more than 60% of bird species. Under the assumption that orders were roughly equivalent in age, this was considered one of the most striking examples of uneven diversity among bird groups. It turns out that this unevenness among groups, though unexpected in some simple models of diversification (the building up of species numbers and morphological variation), is ubiquitous in living organisms, and has profound implications for understanding the broad-scale patterns of evolution that we call macroevolution. For example, one species of shrub in New Caledonia (Amborella trichopoda) is the sister to all other flowering plants (~400,000 species), despite these two lineages being of equal age! Either something has kept Amborella from evolving into new species, or something has accelerated species formation in other flowering plants; alternatively, extinction may have impacted these groups differently. Identifying factors that constrain or promote diversification and extinction is a key goal of macroevolutionary studies.

Starting in the late 1980’s, new phylogenies (evolutionary trees) of birds based on molecular data demonstrated that avian orders were NOT of approximately equal ages. Even so, reanalysis of avian diversity patterns in light of these new trees showed that passerines, especially one group of passerines called the oscines, were still much more diverse than other avian lineages of similar age. This was particularly intriguing because oscines are one of three lineages of birds known to learn their vocalizations, which might contribute to rapid species formation. However, these studies left out many details of relationships within passerines, and in some cases turned out to be just plain wrong. This limited our understanding of evolutionary patterns within this fascinating group.

The widespread commercialization of the polymerase chain reaction (PCR, invented in 1983), combined with Sanger sequencing technology (invented almost a decade earlier than PCR),  added fuel to the ongoing revolution in our understanding of bird (and passerine) relationships. One of the most interesting results uncovered in studies of passerines was the idea that the group had its origin in Gondwana (the southern supercontinent that subsequently broke up into South America, Africa, India, Australia, and New Zealand). This was motivated by the observation that major passerine groups were ancestrally Gondwanan, including the New World suboscines (ancestrally South American), and unexpectedly the oscines (ancestrally Australasian). A controversial implication of this work was that passerines dated back to before the Cretaceous/Paleogene boundary, the crucial period in Earth’s history where the non-avian dinosaurs and many other lineages went extinct. Avian paleontologists objected to this assertion, and some subsequent molecular studies that explicitly included fossil data contradicted it.

Addressing this controversy with an integrated molecular and fossil approach for all passerines was a major goal of our OpenWings collaboration. Our recently published study included an unprecedented sample both of species (137 families) and genetic loci (4,060 loci). Sampling thousands of loci (and millions of base pairs) has been made possible by “next generation” sequencing technologies, making our study “phylogenomic”, a somewhat vague term distinguishing phylogenetic studies of typically thousands of loci from the studies of one or a handful of loci that were common during the 90’s and early oughts. Importantly, our study also integrated the (somewhat fragmentary) fossil record of passerines and their relatives with the molecular data.

The bottom line? There are several. In agreement with paleontological perspectives and some recent molecular studies, passerines do NOT appear to be Cretaceous in age, instead dating to the more recent Eocene around 47 million years ago. So, passerines are still old but not nearly as old as some (including myself) speculated. This implies a more complex history of dispersal and possibly extinction than a simple Gondwanan breakup scenario. Second, there is a lot of variation in diversification rates among passerine lineages. Interestingly, this doesn’t appear to be related either to global changes in temperature or to the ecological opportunities lineages tend to experience when dispersing to new areas (e.g., continents). Although worldwide (except Antarctica) in distribution, passerines don’t seem to have speciated rapidly just because they found new places to live (with some notable exceptions, including Darwin’s finches in the Galapagos, and white-eyes in the southwest Pacific). Instead, some lineages seem to speciate more rapidly than others idiosyncratically, in ways that might relate to things like dispersal potential (which in other studies has been shown to slow down speciation in some South American suboscines), dietary innovation (our study suggests that one oscine lineage that exploits sugars and starches seems to have rapidly diversified), and possibly other factors such as reduced evolutionary constraint (i.e. the ability to rapidly evolve diverse feeding morphologies seen in Darwin’s finches).

Testing some of these ideas will depend on future work with more extensive sampling of species. This study only included 137 lineages of nearly 6000 passerines, so there is a lot of work left to do. Some of our collaborators (and others) are already preparing a species-level phylogeny of suboscine passerines (over 1000 species), and other projects are proceeding apace. Also, placing passerines in the broader context of a new fossil-calibrated phylogenomic tree of birds will more reliably place their diversity in the broader context of avian diversity. Stay tuned for the next chapter!