Darren Naish in a recent post on his most excellent Tetrapod Zoology blog on a completely different subject mentioned the giant fossil bird Eremopezus, which inspired me to look it up (I was nearly inspired to change subject by watching a bagmoth in the lab here sealing itself into its bag in preparation for pupating, but another time, perhaps...)
Eremopezus is known from leg bones from the upper Eocene of the Fayum of Egypt. The most recent review is by Rasmussen et al. (2001), but it was first described in 1904. Lambrecht later divided the then-available material into two genera, Eremopezus Andrews 1904 and Stromeria Lambrecht 1929, but there is little significant difference between material assigned to the two and they are now regarded as synonymous.
Being a giant landbird, Eremopezus was originally thought to be related to modern giant landbirds, the ratites. Ratites are a group of flightless birds distributed between the southern continents - the ostrich (Africa), rheas (South America), emu, cassowaries (Australia), moa and kiwis (New Zealand). Arguments have run back and forth about whether the ratites are monophyletic, or have arisen independently from different ancestors. Recent molecular phylogenies have been pretty much unanimous that the ratites are indeed monophyletic, and together with the flighted tinamous (Tinamidae) are the sister group to the remaining modern birds. On the basis of a prominent ridge on the tarsometatarsus, Lambrecht (1933) suggested that Eremopezus was related to the elephant birds (Aepyornithidae) of Madagascar.
The problem is that this is simply not very significant evidence, as pointed out by Rasmussen et al. (2001). Large flightless birds show a great deal of similarity in the form of the leg bones, due to similar functional requirements. The flightless carnivorous bird Diatryma has hindlimb bones indistinguishable from those of ratites, despite being more closely related to the modern Anseriformes (ducks and geese). Rasmussen et al. concluded that Eremopezus could not be reliably associated with any other known group of birds.
To add to this, Eremopezus showed a number of distinct features all of its own. It appears to have been a fairly lightly-built bird, but slightly larger than a cassowary or rhea. The distal end of the tarsometatarsus is markedly flattened dorsoventrally, and the trochleae (and hence the toes) are quite widely splayed (an attachment scar indicating the presence of a hallux - the rear-pointing toe - is present, but this was probably small as an adaptation for terrestriality). The trochleae on either side have relatively light grooves, suggesting that the toes were quite mobile. The modern birds with the most similar morphologies are Sagittarius serpentarius (secretarybird) and Balaeniceps rex (shoebill). Both these birds use their feet for manipulation - Sagittarius is a ground predator that catches prey with its feet, while Balaeniceps uses its feet to grasp floating vegetation in swampy habitats. The Fayum of the Eocene also appears to have been a quite swampy habitat, but the appeal of a gigantic secretarybird is not to be denied. In the meantime, we simply have to wait on further remains to turn up before we can say more on the subject.
7 hours ago in The Phytophactor