Journal of Ornithology

, Volume 148, Supplement 2, pp 455–458

The renaissance of avian paleontology and its bearing on the higher-level phylogeny of birds


    • Forschungsinstitut Senckenberg, Sektion Ornithologie

DOI: 10.1007/s10336-007-0159-8

Cite this article as:
Mayr, G. J Ornithol (2007) 148: 455. doi:10.1007/s10336-007-0159-8


Recent phylogenetic analyses provide strong evidence for some previously undetected clades of morphologically very divergent avian groups. Also, within the past decades, palaeornithology has experienced a renaissance and the Paleogene fossil record of birds is approaching that of mammals in the number of recorded higher-level taxa. However, there is still little mutual exchange between students of these different data, as molecular systematists are often unfamiliar with the fossil record of birds, whereas palaeornithologists only recently started to interpret their fossils in the light of modern phylogenetic analyses. Here, this deficiency is remedied in a brief overview of some fossil “missing links” between extant higher avian groups, which combine derived characters of a certain taxon and plesiomorphic characters of its sister group.


Fossil birdsMissing linksPalaeornithologyPhylogeny


In the past few decades, the Paleogene (pre-Miocene) fossil record of birds has approached that of mammals, at least in terms of the recognized higher-level taxa (e.g., Mayr 2005a). However, the avian fossil record is still overseen by only a few specialists and, unlike the situation in mammalogy (e.g., Novacek 1992; Asher et al. 2003; Boisserie et al. 2005), fossil taxa are usually not considered in analyses of and discussions on the phylogeny of extant avian taxa, other than calibrations of molecular clocks (e.g., Cracraft et al. 2004; Livezey and Zusi 2007). Livezey and Zusi (2007, p. 4) even went as far as to note that fossil birds “typically provide only substandard anatomical material or incomplete specimens”. Certainly, this is not true. Many Paleogene taxa are known from complete skeletons which are often complemented by three-dimensionally preserved isolated bones, and sometimes even exhibit remarkable soft-tissue preservation (e.g., Mayr 2005a, 2006).

The significance of fossils as “missing links” (transitional forms) between morphologically disparate clades of extant birds has been vilified as the “most prevalent idiosyncracy [sic] of paleontological perspectives” by Livezey and Zusi (2007, p. 49). Indeed, fossils are not mandatory for a reconstruction of the phylogeny of extant organisms. However, stem group representatives of a certain clade may support existing phylogenetic hypotheses on widely divergent taxa, if they exhibit old bauplan characteristics which were reduced or transformed in the crown group. In the following, three examples of such “missing links” are given, which provide independent evidence for recently established phylogenetic hypotheses.

Flamingos (Phoenicopteriformes) and grebes (Podicipediformes)

Sister group relationship between flamingos (Phoenicopteriformes) and grebes (Podicipediformes) was first proposed in a molecular study of van Tuinen et al. (2001), who analyzed mitochondrial and nuclear DNA sequences as well as DNA–DNA hybridization data, and also resulted from sequence analyses of the nuclear ZENK gene (Chubb 2004), the RAG-2 exon (Cracraft et al. 2004), RAG-1 gene (Ericson et al. 2006: Fig. ESM 2), the myoglobin intron 2 (Ericson et al. 2006: Fig. ESM 3), and combined sequences of ADH5, GPD3–5, FGB7, 12S rDNA, 16S rDNA, and tRNA Valine (Fain and Houde 2007). Derived morphological characters shared by flamingos and grebes were first identified by Mayr (2004) and augmented by Manegold (2006).

A flamingo–grebe clade is strongly supported by the fossil Palaelodidae, which have an abundant fossil record in the Paleogene and Neogene of Europe (Mayr 2004, 2005a; note that Livezey and Zusi 2007, p. 50, misquoted Mayr (2005a) in stating that the “intermediacy of two skeletal features between Juncitarsus […] and the Palaelodidae […] provides a morphological link between Phoenicopteriformes and Podicipediformes”; the comments in that study pertained to Palaelodus only). Palaelodids are unanimously considered stem lineage representatives of the Phoenicopteriformes and possess a deep lower jaw which suggests the existence of a “primitive filter-feeding apparatus” (Cheneval and Escuillié 1992, p. 209). However, they have much shorter hindlimbs than crown group Phoenicopteriformes and the tarsometatarsi in particular “show many similarities with those of a foot-propelled diving bird such as Podiceps [Podicipedidae]” (Cheneval and Escuillié 1992). In the light of the new phylogenetic hypotheses, it is reasonable to consider the similarities between Palaelodidae and Podicipedidae to be plesiomorphic. Because grebes and palaelodids are aquatic birds which use their hindlimbs for propulsion, it is thus most parsimonious to assume that the stem species of the Phoenicopteriformes was also an aquatic bird which used its hind limbs for propulsion in the water (Mayr 2004; contra Storer 2006). Species on the stem lineage of the Phoenicopteridae then entered a new ecological zone, as filter feeders in shallow waters (see Mayr 2007a).

Buttonquails (Turnicidae) and shorebirds (Charadriiformes)

Molecular analyses provide strong evidence that buttonquails (Turnicidae), traditionally most often considered aberrant “gruiform birds”, are representatives of the Charadriiformes (shorebirds and allies) (Paton et al. 2003; Cracraft et al. 2004; Fain and Houde 2004, 2007; Paton and Baker 2006; Ericson et al. 2006). Derived morphological features which are shared by crown group Turnicidae and other charadriiform birds include a derived shape of the extremitas sternalis of the coracoid which forms three pointed projections, the absence of pneumatic foramina in the fossa pneumotricipitalis of the humerus, the presence of a tuberculum on the os carpi ulnare at the insertion area of the ligamentum humerocarpale, and the fact that the fourth phalanx of the fourth toe is shorter than the third phalanx (Mayr and Knopf 2007). Charadriiform affinities of the Turnicidae are further supported by the osteology of the early Oligocene fossil taxon Turnipax, which is known from skeletons from Céreste in France and Frauenweiler in Germany. The two species of Turnipax are small birds and combine derived characteristics of crown group Turnicidae (including a highly peculiar morphology of the coracoid) with a plesiomorphic “charadriiform” overall morphology (Mayr 2000, Mayr and Knopf 2007). Before any molecular evidence for charadriiform affinities of the Turnicidae was available, the striking character mosaic of this fossil taxon raised the question whether “buttonquails are highly modified Charadriiformes which evolved from a Turnipax-like ancestor” (Mayr 2000, p. 629).

Owlet-nightjars (Aegothelidae) and apodiform birds

It has recently also been suggested that the “Caprimulgiformes” (nightjars and allies) are not monophyletic, and that owlet-nightjars (Aegothelidae) are the sister group of swifts and hummingbirds (Apodiformes). The owlet-nightjar–apodiform clade was first proposed from an analysis of morphological characters, and apomorphies of this clade include a derived morphology of the quadrate and the splenius capitis muscle (Mayr 2002a). A clade (Aegothelidae + Apodiformes) is supported by an indel in the c-myc gene (Mayr et al. 2003) and resulted from all subsequent molecular analyses (Cracraft et al. 2004; Ericson et al. 2006; Barrowclough et al. 2006).

Nevertheless, the extant representatives of this clade are morphologically very different, and owlet-nightjars, for example, differ markedly from apodiform birds in their feathering.

In the light of the novel phylogenetic hypothesis, it is thus notable that the Middle Eocene Parargornis messelensis, a stem group representative of the Trochilidae, exhibits an owlet-nightjar-like feathering with short and rounded wings and a long tail (Mayr 2005b). Parargornis messelensis further has an owlet-nightjar or swift-like beak and certainly was an insectivorous bird. It is most parsimonious to assume that the feathering of Parargornis is plesiomorphic for the total group of Trochilidae (stem group + crown group); the narrow wings of the crown group representatives are functionally correlated with the highly specialized hovering flight of the nectarivorous crown group representatives.


All of the clades discussed above: (1) can be established with morphological apomorphies found in the crown group representatives; (2) result from independent analyses of a fair number of different gene sequences; and (3) are further supported by the mosaic character distribution of fossil avian taxa. Remarkably, however, none is recovered in the analysis of Livezey and Zusi (2007) which was based on 2,954 characters—probably the largest data set of morphological characters analyzed so far. A discussion of this analysis is beyond the scope of the present paper (see Mayr 2007b). However, the authors included hardly any fossil taxa in their study, and most of those that were considered are nested within extant clades of lower taxonomic rank (e.g., the columbiform dodo) and thus do not contribute to a resolution of the higher-level phylogeny of neornithine birds. Omission of fossils from Livezey and Zusi’s (2007) analysis is so much the more regrettable, as it has been shown that even very incompletely known fossil taxa increase the accuracy of phylogenetic analyses (e.g., Santini and Tyler 2004).

Many Paleogene taxa exhibit a completely unexpected mosaic combination of characters found in extant birds traditionally considered unrelated (e.g., Mayr 1999, 2001, 2002b). Although phylogenetic interpretations of fossils with an all-too suggestive character combination can of course be misleading (see, e.g., Livezey and Zusi 2007, p. 49, concerning the case of the long-legged anseriform Presbyornis), fossil taxa may also be consulted a posteriori to make sense of well-supported hypotheses based on molecular data which group together morphologically disparate extant taxa. Eventually, fossils are the only direct evidence of the ancient diversity of life, and it is about time that avian systematists raise their awareness of this often neglected source of phylogenetic information.


Neuere phylogenetische Analysen stützen bisher nicht erkannte Monophyla, welche morphologisch sehr unterschiedliche rezente Vogelgruppen beinhalten. Innerhalb der letzten Jahrzehnte hat die Paläornithologie zudem eine Renaissance erfahren und der paläogene Fossilbericht der Vögel nähert sich bezüglich der Anzahl der bekannten höherrangigen Taxa demjenigen der Säugetiere an. Es gibt jedoch noch wenig gegenseitigen Austausch zwischen den Bearbeitern dieser unterschiedlichen Daten, da Molekularsystematiker häufig mit dem Fossilbericht der Vögel nicht vertraut sind, während Paläornithologen gerade erst beginnen, ihre Fossilien im Hinblick auf moderne phylogenetischen Analysen zu deuten. Hier werden einige fossile Gruppen vorgestellt, die als „missing links” (morphologische Zwischenformen) zwischen heutigen höherrangigen Vogeltaxa betrachtet werden können, und abgeleitete Merkmale eines Taxons mit plesiomorphen Merkmale seiner Schwestergruppe kombinieren.


I thank two anonymous referees for comments which improved the manuscript.

Copyright information

© Dt. Ornithologen-Gesellschaft e.V. 2007