No unambiguously paleognathous fossil birds are known until the
Cenozoic (though birds occasionally interpreted as
lithornithids occur in
Albian appalachian sites), but there have been many reports of putative paleognaths, and it has long been inferred that they may have evolved in the
Cretaceous. Given the Northern Hemisphere location of the morphologically most
basal fossil forms (such as
Lithornis,
Pseudocrypturus,
Paracathartes and
Palaeotis), a
Laurasian origin for the group can be inferred. The present almost entirely
Gondwanan distribution would then have resulted from multiple colonisations of the southern landmasses by flying forms that subsequently evolved flightlessness, and in many cases, gigantism. '' fossil cast,
Copenhagen Zoological Museum One study of molecular and paleontological data found that modern bird orders, including the paleognathous ones, began diverging from one another in the
Early Cretaceous.
Benton (2005) summarized this and other molecular studies as implying that paleognaths should have arisen 110 to 120 million years ago in the Early Cretaceous. He points out, however, that there is no fossil record until 70 million years ago, leaving a 45 million year gap. He asks whether the paleognath fossils will be found one day, or whether the estimated rates of
molecular evolution are too slow, and that bird evolution actually accelerated during an adaptive radiation after the
Cretaceous–Paleogene boundary (K–Pg boundary). Before the advent of genetic analysis, some authors questioned the
monophyly of the Palaeognathae on various grounds, suggesting that they could be a hodgepodge of unrelated birds that have come to be grouped together because they are coincidentally flightless. Unrelated birds might have developed ratite-like anatomies multiple times around the world through
convergent evolution. McDowell (1948) asserted that the similarities in the palate anatomy of paleognaths might actually be
neoteny, or retained embryonic features. He noted that there were other features of the skull, such as the retention of sutures into adulthood, that were like those of juvenile birds. Thus, perhaps the characteristic palate was actually a frozen stage that many carinate bird embryos passed through during development. The retention of early developmental stages, then, may have been a mechanism by which various birds became flightless and came to look similar to one another. ''. Hope (2002) reviewed all known bird fossils from the Mesozoic looking for evidence of the origin of the evolutionary radiation of the
Neornithes. That radiation would also signal that the paleognaths had already diverged. She notes five
Early Cretaceous taxa that have been assigned to the Palaeognathae. She finds that none of them can be clearly assigned as such. However, she does find evidence that the Neognathae and, therefore, also the Palaeognathae had diverged no later than the Early
Campanian age of the
Cretaceous period.
Vegavis is a fossil bird from the
Maastrichtian stage of
Late Cretaceous Antarctica.
Vegavis is most closely related to true ducks. Because virtually all phylogenetic analyses predict that ducks diverged after paleognaths, this is evidence that paleognaths had already arisen well before that time. An exceptionally preserved specimen of the extinct flying paleognathe
Lithornis was published by Leonard et al. in 2005. It is an articulated and nearly complete fossil from the early
Eocene of Denmark, and thought to have the best preserved lithornithiform skull ever found. The authors concluded that
Lithornis was a close
sister taxon to tinamous, rather than ostriches, and that the lithornithiforms + tinamous were sister to the other paleognaths. They concluded that all ratites, therefore, were monophyletic, descending from one common ancestor that became flightless. They also interpret the paleognath-like
Limenavis, from
late Cretaceous Patagonia, as possible evidence of a Cretaceous and
monophyletic origin for paleognaths. An ambitious genomic analysis of the living birds was performed in 2007, and it contradicted Leonard et al. (2005). It found that tinamous are not primitive within the paleognaths, but among the most advanced. This requires multiple events of flightlessness within the paleognaths and partially refutes the Gondwana vicariance hypothesis (see below). The study looked at DNA sequences from 19 loci in 169 species. It recovered evidence that the paleognaths are one natural group (
monophyletic), and that their divergence from other birds is the oldest divergence of any extant bird groups. It also placed the tinamous within the ratites, more derived than ostriches, or rheas and as a sister group to emus and kiwis, and this makes ratites
paraphyletic. A related study addressed the issue of paleognath phylogeny exclusively. It used molecular analysis and looked at twenty unlinked nuclear genes. This study concluded that there were at least three events of flightlessness that produced the different ratite orders, that the similarities between the ratite orders are partly due to
convergent evolution, and that the Palaeognathae are
monophyletic, but the ratites are not. Beginning in 2010, DNA analysis studies have shown that tinamous are the
sister group to extinct
moa of New Zealand. A 2020 molecular study of all bird orders found paleognaths and neognaths to have diverged in the
Late Cretaceous or earlier, before 70 million years ago. However, all modern paleognath orders only originated in the latest
Paleocene and afterwards, with ostriches diverging in the latest Paleocene, rheas in the early
Eocene,
kiwis (and presumably
elephant birds) very shortly after in the early Eocene, and finally
Casuariiformes and
tinamous (and presumably
moas) diverging from one another in the mid-Eocene.
History of classifications In the history of
biology there have been many competing taxonomies of the birds now included in the Palaeognathae. The topic has been studied by
Dubois (1891),
Sharpe (1891),
Shufeldt (1904),
Sibley and
Ahlquist (1972, 1981) and
Cracraft (1981).
Merrem (1813) is often credited with classifying the paleognaths together, and he coined the taxon "Ratitae" (see above). However,
Linnaeus (1758) placed cassowaries, emus, ostriches, and rheas together in
Struthio.
Lesson (1831) added the kiwis to the Ratitae.
Parker (1864) reported the similarities of the palates of the tinamous and ratites, but
Huxley (1867) is more widely credited with this insight. Huxley still placed the tinamous with the Carinatae of Merrem because of their keeled
sterna, and thought that they were most closely related to the
Galliformes. Pycraft (1900) presented a major advance when he coined the term Palaeognathae. He rejected the Ratitae-Carinatae classification that separated tinamous and ratites. He reasoned that a keelless, or "ratite", sternum could easily evolve in unrelated birds that independently became flightless. He also recognized that the ratites were secondarily flightless. His subdivisions were based on the characters of the palatal skeleton and other organ systems. He established seven roughly modern orders of living and fossil paleognaths (Casuarii, Struthiones, Rheae, Dinornithes, Aepyornithes, Apteryges, and Crypturi – the latter his term for tinamous, after the Tinamou genus
Crypturellus). The Palaeognathae are usually considered a
superorder, but authors have treated them as a taxon as high as
subclass (Stresemann 1927–1934) or as low as an
order (Cracraft 1981 and the
IUCN, which includes all paleognaths in an expanded Struthioniformes). Palaeognathae was defined in the
PhyloCode by George Sangster and colleagues in 2022 as "the least inclusive crown clade containing
Tinamus major and
Struthio camelus". Yuri
et al. (2013) named the clades
Notopalaeognathae and
Novaeratitae, the former defined in the
PhyloCode by Sangster
et al. (2022) as "the least inclusive crown clade containing
Rhea americana,
Tinamus major, and
Apteryx australis", while the latter also defined in the
PhyloCode by Sangster
et al. (2022) as "the least inclusive crown clade containing
Apteryx australis and
Casuarius casuarius". Cloutier, A.
et al. (2019) in their molecular study places ostriches as sister to other palaeognaths, with the rhea as the sister lineage to the remaining, non-ostrich palaeognaths. An alternative phylogeny was found by Kuhl, H.
et al. (2020). In this treatment, all members of Palaeognathae are classified in Struthioniformes, but they are still shown as distinct orders here. This lineage containing the sister relationship between tinamous and moas was given the clade name
Dinocrypturi, being named and defined in the
PhyloCode by Sangster
et al. (2022) as "the smallest clade containing
Tinamus major and
Dinornis novaezealandiae". == Description ==