There seem to be three major branches in Coronaves. One is primarily waterbirds (Pelecanae), one includes the shorebirds (Charadriae), and one consisting of land birds (Passerae).
Pelecanae: Waterbirds, Waders, and Cuckoos
The Pelecanae consist of the mostly aquatic and semi-aquatic bird families that are in Aequornithes together with the hoatzin, cranes and rails, cuckoos, and turacos. I used the old name ‘Natatores’ for similar groupings in early versions of this list. Natatores was a name used in the 19th century for waterbirds; although it has been attributed to Baird 1858, its ornithological use substantially predates him. The terms Conglomerati and Cracrafti have also been used for somewhat more inclusive groupings that also include raptors and shorebirds (e.g., Slack et al., 2007).
Two of the troublesome taxa, the hoatzin and cuckoos, are included in the Pelecanae by Hackett et al. (2008). Ericson et al. (2006) also includes the cuckoos, but puts the hoatzin in Metaves. Wang et al.'s results (2011) also suggest that the cuckoos belong in this clade. However, there are many studies placing them elsewhere. Hackett et al. found low support for including the hoatzin, so we shouldn't be surprised if further research moves it around.
Nonetheless, we follow the Hackett tree, where the hoatzin is basal in the Pelecanae. It goes in a separate superorder, Opisthocomimorphae. The remainder then splits into two parts. The first contains the bustards, cuckoos, cranes and rails (Cuculimorphae). The other contains turacos (Musophagimorphae) and seabirds and waders (Aequornithes). Note that the hoatzin has sometimes been considered allied with cuckoos or turacos, so its placement here is not unreasonable (even though I have low confidence in it).
The waterbird clade, Aequornithes (Mayr, 2011), is an assemblage of water and wading birds. This includes most pelagic species (except tropicbirds) and the large waders (but not sunbittern or flamingos, which are in Metaves). I break this group into several orders because there is a substantial chance that more rearrangement will be needed in the future. This group is the core of what Sibley and Monroe called the Ciconiiformes. Here, the Ciconiiformes have been completely dismantled, with only Ciconiidae (storks) remaining. The rest of the grouping consists of the Gaviiformes (loons), Sphenisciformes (penguins), Procellariiformes (seabirds), Suliformes (frigatebirds, boobies, anhingas, and cormorants), Pelecaniformes (hamerkop, shoebill, and pelicans), Plataleiformes (ibises and spoonbills) and Ardeiformes (herons). We have high confidence in the overall grouping, although some uncertainty remains concerning its inner structure.
There are also taxonomic issues in the families inside the Pelecaniformes. The Boat-billed Heron was previously considered to be the only member of the Cochlearidae. The status of two other monotypic families, the Shoebill and Hamerkop, has also been a perennial issue. The analyses of Ericson et al. (2006a) and Hackett et al. (2008) indicate that both are relatives of the pelicans. Indeed, the tree allows them to be lumped into the same family. We keep them separate not only because of their uniqueness, but also because the division between them seems to be ancient.
The Gruiformes are another area of major change, with three families lost to Metaves, one lost to Charadriiformes, one lost to Cariamiformes, and one (Otididae, bustards) becoming its own order near the Gruiformes. The flufftails (Sarothruridae) are separated from the Rallidae as they seem more closely related to the finfoots. The Musophagidae (turacos) are now considered sister to the Aequornithes.
Charadriae: Shorebirds, Alcids, Skuas, and Gulls
There is a lot of evidence for the Charadriae grouping (e.g., Ericson et al., 2003a; Paton et al., 2003; Cracraft et al., 2004). Other than arguments about the Herring Gull complex and where the terns and skimmers fit, the taxonomy of this clade is now pretty well worked out. Many studies have found that gulls and alcids are closely related to the shorebirds. A recent development is the recognition that the buttonquails are part of it (Paton et al., 2003; Paton and Baker, 2006; Fain and Houde, 2007). The differences from the Sibley-Monroe list are the loss of the sandgrouse (to Metaves), the addition of the buttonquail, and some reordering of the families. If Metaves turns out to be illusory, it is possible that the sandgrouse (and maybe pigeons) will end up here, as may the Mirandornithes.
Passerae: Higher Land Birds
The third branch of Coronaves is Passerae. It is probably best thought of consisting of three pieces: Accipitrimorphae (hawks and American vultures), ‘Anomalogonatae’ (the “higher land birds”), and the Passerimorphae (seriemas, falcons, parrots, and passerines). I originally had only mediocre confidence in the this grouping as it is not supported in the mitochondrial analyses. However, the appearance of Wang et al. (2011) and especially Suh et al. (2011) leaves me reasonably happy with the grouping of these taxa.
A notable feature of this taxonomy is that the American Vultures are classified in their own order, Cathartiformes (AOU's SACC has also adopted this classification). I have done this partly to reflect the considerable uncertainty about where they fit in. Contrary to Sibley-Monroe and 1990's-style checklists, they are not closely related to storks (see Cracraft et al., 2004; Gibb et al., 2007; Slack et al., 2007). Keep in mind that the vultures also appear no more related to the Accipitriformes than they are to anything else in this superorder.
I have followed Hackett et al. (2008) concerning the placement of the hawks, falcons, and American vultures (Accipitriformes, Falconiformes, and Cathartiformes). However, this is controversial and several alternatives tree should be considered. Slack et al. (2007) put all three orders in Pelecanae. Gibb et al's (2007) results are more ambiguous. They present a tree showing the all three in a group that includes other members of both parts of Coronaves. They also present a network diagram that could be consistent with placing the hawks and falcons between Pelecanae and Passerae. The fact that Gibb et al. include more birds from Passerae (an owl and parrot) in their analysis than Slack et al. did may account for this.
One difference from recent AOU lists is the treatment of the Falconiformes, Accipitriformes, and Cathartiformes. Also, the Cariamidae (seriemas) are moved from the Gruiformes to the Falconiformes. The “terror birds” of ancient South America (Phorusrhacidae) are thought to be related to the seriemas (Alvarenga and Höfling, 2003).
The original Anomalogonatae were named by Garrod (1874), with the Strigiformes added later by Beddard (1898). Most of them have remained together in most taxonomic lists since then. One important defining character was the lack of an ambiens muscle (also lost by some unrelated birds). The ‘true Anomalogonatae’ are Beddard's Anomalogonatae, minus the Passeriformes and Strisores. This leaves a core group consisting of the Leptosomiformes, Strigiformes, Coliiformes, Trogoniformes, Bucerotiformes, Coraciiformes, and Piciformes. Ericson et al. (2006a) and Hackett et al. (2008) found it a monophyletic group, a group that includes a common ancestor and all descendants. Although the details may differ, many of the other analyses find broadly similar groupings. The Coliiformes are probably the most suspect member (Suh et al., 2011), but not so suspect that they fall outside Passerae.
Other than their ordering and placement in the ‘Anomalogonatae’, the treatment of much of the remaining non-passerine families is close to that of the Sibley-Monroe list. Changes over time have mostly involved whether to consider certain groups families or sub-families. One interesting case is the Cuckoo Roller. It was originally considered a cuckoo, some affinities with the rollers were noted, and it has more recently been considered its own family, Leptosomidae. Although it belongs in ‘Anomalogonatae’, its exact position is still unsettled.
The owls, mousebirds, trogons are also placed in separate orders. The hornbills, which are split into Bucorvidae (ground-hornbills) and Bucerotidae (hornbills) form a grouping with the hoopoes and woodhoopoes. The rollers, bee-eaters, todies, motmots, and kingfishers form the Coraciiformes. The consensus seems to be to leave the three kingfisher subfamilies as subfamilies, and I follow that. That brings us to the Piciformes.
The classification of the Piciformes follows the AOU's South American Classification Committee rather than the AOU checklist. This means the Capitonidae (American barbets), Semnornithidae (toucan-barbets), and Ramphastidae (toucans) have family status and that requires the Asian and African barbets each have their own family (Moyle, 2004; Johansson and Ericson, 2003). All of the new world forms are more closely related to each other than to the old world barbets. The tree makes this clear.
Passerimorphae: Seriemas, Falcons, Parrots, and Passerines
This brings us to the Passerimorphae, which includes the seriemas, falcons, parrots, and passerines. This clade was one the radical suggestions of Ericson et al. (2006), and subsequently found by Hackett et al. (2008). Although it is sometimes thought that the controversial β-fibrinogen gene creates this grouping, that is not so. It appears on Ericson et al.'s ESM-6 tree that explicitly excludes β-fibrinogen.
Several newer papers provide additional support for this clade. Nabholz et al. (2011) suggested that mitochondrial studies that seemed to contradict this notion were artifacts produced by changes in base composition, and that mitochondrial evidence actually supported placing the parrots sister to the passerines.
Suh et al. (2011), using retroposons, found support for most of the Passerae clade, for the narrower Passerimorphae, and for a sister relationship between the parrots and passerines. They proposed the name Eufalconimorphae for Passerimorphae and Psittacopasserae for the parrot/passerine clade. Wang et al. (2011) using a new, independent dataset of 30 nuclear loci, found strong support for both groups, although it was not entirely clear whether the seriemas are sister to the falcons or form a basal branch in Passerimorphae. McCormack et al. (2013) focused on ultraconserved elements. They also found support for this clade, although their tree surprisingly shows the Coliidae as its sister group. Given this, I have high confidence in both the Passerimorphae and Psittacopasserae. The papers by Suh et al. (2011), Wang et al. (2011), and McCormack et al. (2013) also strengthened my confidence in the Passerae clade.
Passeriformes: The Songbirds
All of the non-passerines together amount to about 4000 species. The rest of the birds, almost 6000 species, are Passerines, songbirds. They are all members of a single order—Passeriformes.
Although we have long known which birds are passerines and which are not, their relationships have been poorly understood. A comparison of Clements 5th edition (which uses an old taxonomy) and Howard-Moore 3rd edition (more recent, but not current) shows how much revision has been necessary. There are still many passerines that are classified in the wrong family (and genus) which makes it harder to determine proper family boundaries and relations. Recent work on passerine taxonomy has done much to clarify the situation, and I've interspersed discussion of that work in passerine portion of the annotated checklist.
The overall organization of the Passeriformes is fairly well understood now. After a basal split between the New Zealand wrens and all other songbirds (Barker et al., 2002; Barker et al., 2004), the big division is between the suboscines and the oscines. Like the passerines as a whole, the suboscines have generally been identifiable as suboscine, but teasing out the relationships between the suboscines has been difficult. The suboscine branch is comprised of two groups, an old world group (Eurylaimides) and a new world group (Tyrannides). The latter contains the ovenbirds (Furnariida) and a group consisting of the tyrant flycatchers and allies (Tyrannida).
Most of the passerines are in the oscines. Sibley and Ahlquist's view was that the remaining passerines split cleanly into a corvid group and a group containing everything else (Passerida). Further study of these families has shown that reality is more complex. Unlike the Passerida, the SAM corvids are not a monophyletic group. A number of groups split off separately before we come to the big split between the Corvida and Passerida. There are also some smaller groups in the Passerida, but it mostly splits into 3-5 major groups. I settled on 5 for these web pages. The groups are: Paroidea (tits and fairy-flycatchers) Sylvioidea (babblers and Old World warblers), Certhioidea (creepers, nuthatches, wrens), Muscicapoidea (thrushes, chats, starlings), and Passeroidea (finches, sparrows, tanagers).