It’s a measure of how unreliable family trees are when they’re calculated using just bone shapes, that the method never revealed that the parrot and sparrow groups were sisters. Gerald Mayr’s new paper on the two groups and their feet, explicitly mentions that it took DNA analysis for this relationship to appear. However, trees based on bone shapes might not be all bad – let’s hope not, since he now uses them for working out where exactly some fossil birds, somewhere between the gigantic sparrow-based group (Passeriformes) and the parrot (Psittaciformes) group, fitted in.
But what about their feet, you will ask? Sparrows and allies (Passeriformes) have ordinary toes: the inner one, equivalent to our big toe, pointing back, and the other three forward. But parrots are zygodactylic – they point the outer one, toe 4, back too.
When did all that happen? Falcons, the closest living relatives and likely ancestors of the sparrows/parrots groupage, are more normal, but some other birds that are also descended from birds of prey, such as woodpeckers, also point two toes back. (So do owls, but it’s looking like this happened several times, and seems to be an adaptation to perching on, and maybe nesting within, vertical tree trunks, instead of in nests on branches.) Mayr explains in this paper, that there was a large group of zygodactylic birds, the Zygodactylidae, that were in fact a closer sister group to the Passeriformes than the parrots are. “…[T]he distal end of the tarsometatarsus of zygodactylids is reminiscent of that of parrots […]. The rest of the skeleton of these long-legged birds is, however, very similar to that of passerines and shows little resemblance to parrots.” (Long-legged and zygodactylic? Why?! Dunno.)
Mayr names a new clade holding just Zygodactylidae and passerines: “A new name, Parapasseres, is further introduced for the clade including zygodactylids and passerines but not Psittacopes (Fig. 7A). This clade is diagnosed by the presence of a large processus intermetacarpalis of the carpometacarpus and by a long and slender tarsometatarsus, which distinctly exceeds the humerus in length and exhibits a crista plantaris lateralis.”
He also places these three birds into their own unnamed clade, sister to his new Parapasseres (=Zygodactylidae + passerines):
(In the paper he also offers two furthertrees, one forcing monophyletic Psittacopasseres – parrots and passerines, but with Mayr’s corrected plural spelling, and one forcing monophyletic Eufalconimorphae – the group named by Alex Suh et al. in 2011 to contain falcons parrots and passerines. In the tree shown here near the top, the falcons – Falconidae – are much closer to the owl group Strigidae than to parrots – Psittacidae – just above the middle and Passeriformes, in the top of the grey box. Molecular trees suggest that isn’t right, but if the fossils are over 50my old you have only the bones, and with luck the grey box may be right.)
Mayr explains an idea where “…the formation of zygodactyl feet in parrots is due to a disproportionate development of the abductor and extensor muscles of the fourth toe, with the musculus abductor digiti IV being strongly developed and the musculus extensor brevis digiti IV greatly reduced. It is this asymmetric action of the muscles, which is believed to trigger the formation of the osteological structures associated with the zygodactyl foot in early osteogenesis. In passerines, both muscles, as well as all other intrinsic foot muscles, are greatly fossil or completely lost. This restores the balance of the muscular forces acting on the fourth toe and prevents the toe from turning backwards in early ontogeny, therefore invalidating the epigenetic mechanisms, which cause the complex derived shape of the tarsometatarsal trochlea of this toe in the zygodactyl parrots”
Unfortunately he then goes on to say that since the Zygodactylidae, that sister group to passerines, appears to have had a well-developed musculus extensor brevis digiti IV, that can’t explain why their 4th toe points backwards. Mind you, nature often has a way of taking one feature, say a backward 4th toe caused by one reason, and changing the way it’s implemented but retaining the feature.
Anyway, he’s put some trees together (necessarily based on bone shapes but often forcing groupings taken from DNA analyses), where passerines do seem to be descended from zygodactylic ancestors. Which also lived in hollow trees. As did the ancestors of woodpeckers. And hornbills, toucans, and hoopoes still do, and they all seem to have descended from birds of prey too.