Evolution and Systematics
Discussion of Phylogenetic Relationships
Relationships among paleognaths have been controversial and are likely to remain so for some time, but the tree shown above is the current best estimate, based on analysis of 20 unlinked nuclear genes (Harshman et al. 2008). This tree shows the ratites (flightless paleognaths, including all extant paleognaths except the tinamous) to be polyphyletic, a major departure from most recent analyses of molecular (Cooper et al. 2001, Haddrath and Baker 2001, Sibley and Ahlquist 1990) and morphological (Lee et al. 1997, Livezey and Zusi 2007) data.
All possible relationships among the paleognaths, including complete polyphyly, have been proposed at various times (Olson 1985, Houde 1988), but in the past few decades, a few points have been settled. Paleognath monophyly has become clear, and it is universally agreed that Dromaiidae (emus) and Casuaridae (cassowaries) are close relatives. And, until recently, monophyly of the ratites had been generally accepted. Beyond this, however, there has been no agreement. Most morphological analyses have found Apterygidae (kiwis) to be the sister group of all other ratites. Most molecular analyses have found kiwis to be the sister group of emus and cassowaries.
There have been a few holdouts against ratite monophyly; both Elzanowski (1995) and Bock and Buhler (1990), on the basis of morphological characters, found Struthionidae (ostriches) to be the sister group of all other paleognaths, and this is also what the nuclear DNA data show. Nuclear data also find, in accord with previous molecular analyses, that kiwis are the sister group of the emus and cassowaries, the three groups together consituting a clade of Australasian ratites. However, the data are unable to resolve relationships among the Australasian ratites, Rheidae (rheas), and tinamous.
The importance of evolutionary models in phylogenetic analyses is borne out by the most recent analyses of mitochondrial data (Phillips et al., 2010), in which the same data that previously resulted in ratite monophyly have been reanalyzed using a more complex model, with results consistent with the nuclear DNA tree above; that is, with ostrich as the sister group of all other paleognaths, though support is poor.
This tree implies that the various flightless paleognaths lost their ability to fly (and thus their sternal keels) independently of each other. The alternative possibility, that tinamous regained the ability to fly (and their sternal keels) after inheriting flightlessness from a common ratite ancestor, is considered less likely.
Placement of the extinct groups is unclear. Morphological and mitochondrial DNA data are in disgreement on relationships of Dinornithidae (moas), and there are so far no nuclear DNA sequences for extinct birds. The recent reanalysis of mitochondrial data (Phillips et al., 2010) found moas to be the sister group of tinamous. This is also in agreement with a combined analysis of nuclear and mitochondrial data (Harshman, unpublished), in which moas were represented by mitochondrial data only.
For more information about alternative hypotheses see Paleognath Relationships.