The following critiques express the opinions of the individual evaluators regarding the strengths, weaknesses, and value of the books they review. As such, the appraisals are subjective assessments and do not necessarily reflect the opinions of the editors or any official policy of the The American Ornithologists' Union.
Cenozoic Birds of the World, Part 1: Europe.–Jirí Mlíkovsky. 2002. Ninox Press, Prague, 406 pp., ISBN 80-901105-3-8. — This book is a catalogue of the Tertiary birds of Europe; according to the title, it will be followed by other volumes to cover the avifauna of the world. Thus, Jirí Mlíkovsky has undertaken a gigantic task. Previous catalogues included a list of the extinct or recent species recorded as fossils, with their age and typelocality, but Mlíkovsky includes all the localities where a given species has been reported, with references, catalogue numbers, specimen repositories, and so on.
In previous catalogues, few systematic determinations were modified by the authors; Mlíkovsky, however, completely overturns many previous identifications. Attribution of a fossil to a given systematic entity is partly a matter of personal interpretation and subjectivity (and anybody can commit errors), but one has the impression that the guiding principle of Mlíkovsky's work is that everybody before him has been wrong. The result is that this book contradicts practically all that has been written on the Tertiary birds of Europe, which will have serious repercussions for avian paleontology.
The systematics used in the book is a modified version of the one Mlíkovsky presented at the 18th International Ornithological Congress in Moscow in 1982. The system was founded on “a large amount of data…evaluated by means of new methods of biological systematics [developed by Mlíkovsky], which have been based especially on new achievements in mathematical logics” (Mlíkovsky 1985). This classification included four subclasses, the Archaeopterygidae (for the Jurassic radiation), the Hesperornithidae (for the Cretaceous radiation), the Passerida (for the first branch of the Cenozoic radiation), and the Ciconiida (for the second branch of the Cenozoic radiation). To give just one example of this strange classification, the Apterygiformes include, distributed among four suborders, the Procellariidae, Diomedeidae, Hydrobatidae, Pelecanoididae, Spheniscidae, Dinornithidae, Anomalopterygidae, Apterygidae, Tinamidae, Podicipedidae, Dromadidae, and Chionididae. Sibley and Ahlquist (1990: 240), in their chronological survey of classification, dismissed Mlíkovsky's scheme with a simple, “No comment.”
Although based on previous ones, the new systematics used here includes numerous modifications, without explanations. In the Charadriiformes, for example, are found the Jacanidae and Rostratulidae (previously Anseriformes), the Scolopacidae (previously Plataleiformes), the Glareolidae, Pterocletidae, Charadriidae, Laridae (previously Ciconiiformes), and the Alcidae (previously Alciformes).
The classification within superorders is also very anomalous. For example, the Piciformes are in the superorder Ciconii, whereas the Passeriformes are in the superorder Passeri; likewise, the Anseriformes are in the Ciconii, whereas the Galliformes are in the Passeri. The Bucerotiformes, which include three recent families and one extinct family, are placed in the Gavii. This new classification is said to be founded on many different kinds of data—morphological and molecular on the one hand, behavioral and ecological on the other. “This arrangement is supported by an enormous amount of further data, which will be presented elsewhere” (J. Mlíkovsky unpubl. data). But nothing is given here to explain or justify the author's patently bizarre classification.
I have noticed in the past that Mlíkovsky has difficulty in recognizing similarities between the different shapes of bird bones, and consequently, when identifying fossil birds, he often makes errors. For example, in 1999 he described what he believed to be a Miocene jacana, from Czechia, as Nupharanassa bohemica, but this turned out to be a roller-like bird, belonging to the extinct genus Geranopterus (Mlíkovsky 1999, Mourer-Chauviré 1999, Mayr and Mourer-Chauviré 2000). The tarsometatarsus in the Jacanidae is very characteristic in the very wide distal foramen. In addition, the internal and external trochleae are much shorter than the middle trochlea, and the three trochleae are arranged on a strongly curved line. The supposed Czech jacana was different from the recent Jacanidae in all three of the characters. Mlíkovsky is in complete disagreement with the attribution of this fossil to the rollers. He writes: “No points were presented for its exclusion from the Jacanidae” (p. 126), though the three points mentioned above were indeed presented. I have the impression that, when Mlíkovsky looks at a bone, his perception of it is totally different from that of anyone else.
Early Tertiary birds show an extraordinary diversity and sometimes present mosaics of characters that can be extremely different from the combinations found in recent families. In some instances, it is possible to show that fossils that retain certain primitive characters are on a phylogenetic line leading to a recent family. Previous researchers have sometimes created extinct orders, or more often extinct families, for those forms; but Mlíkovsky does not admit the validity of those taxa and places the extinct orders and the near totality of the extinct families either in synonymy with recent ones, or in the Aves incertae sedis, which has the effect of obscuring a large amount of information on the diversity and early evolution of Paleogene birds. In other cases, those families are wrongly placed in orders to which they are unrelated. For example, the Lithornithiformes (p. 58) are placed in the Tinamiformes, although they are morphologically very different; the Messelornithidae (p. 87) are placed in the Ciconiiformes, and come between the Pelecanidae and the Gruidae, although they are related to the Eurypygidae; the Gastornithidae (p. 94) are placed within the Ciconiiformes, after the Threskiornithidae, although their affinities are with the Anseriformes.
All the small zygodactyl or semizygodactyl forms, generally from Eocene deposits, are jumbled together in the family Zygodactylidae (p. 141). Those forms had been assigned to the families Primoscenidae, Sandcoleidae, Quercypsittidae, Miopiconidae, and Pseudasturidae. Some of the fossils were described from complete skeletons (e.g. Middle Eocene of Messel, Germany) or from associated elements (e.g. Early Eocene of Walton-on-the-Naze, England). The affinities of the forms had been carefully studied by their describers, taking into account not only the tarsometatarsus but also the other elements of the skeleton, and it was clear that all the forms could not belong to a single family. Mlíkovsky also includes, in this grab-bag family, genera that had been described as Coliiformes (Primocolius, Masillacolius, Selmes), Psittaciformes (Psittacopes), and Falconiformes (Messelastur).
Arbitrary attributions or synonymizations also occur at the generic level. Here is just one example. The genus Elaphrocnemus (Cariamae, Idiornithidae), very abundant in the Phosphorites du Quercy, is synonymized (p. 179) with the genus Talantatos Reichenbach, 1852, from the Late Eocene of the Paris Basin. This generic name has never been used in any scientific work until Mlíkovsky (1995) rescued it from the oblivion into which it had sunk after 1852. The holotype of Talantatos fossilis is a femur, partially embedded in stone, with the head missing, and the distal part made up of a large number of small pieces stuck back together. However, the morphological characteristics that are still observable are quite different from those of Elaphrocnemus; consequently, this destabilizing synonymization is unjustified.
Another remarkable case is Mlíkovsky's placement of the genus Diatropornis (Eocene or Oligocene of Quercy) in Aves incertae sedis (p. 270), which would negate the presence of Cathartidae in the Early Tertiary of the Old World. Although the systematic position of Diatropornis was at first considered uncertain, Brodkorb (1964) recognized that it was a cathartid, and this assignment has been supported by all subsequent avian palaeontologists, including Mlíkovsky himself (1996:804). His arguments against Diatropornis being a cathartid are either weak or totally unfounded, and do not address the shape of the hypotarsus, which is absolutely characteristic of the Cathartidae. The occurrence of Cathartidae in Europe is further confirmed by the presence of another genus, Parasarcoramphus (Mourer-Chauviré 2002).
A large number of Late Paleogene and Early Neogene birds are assigned by Mlíkovsky to modern genera. For example in his Eocene, Oligocene, and Miocene “Phasianidae,” Mlíkovsky retains only the genera Paraortyx, Pirortyx, and Miogallus; all other fossil genera are synonymized with the recent genera Coturnix, Alectoris, and Pavo (pp. 152–155). Concerning the genera Palaeortyx, Palaeoperdix, Schaubortyx, and Taoperdix, he says that it is not possible to distinguish them morphologically from the recent genus Coturnix. Therefore, he places all the small forms in the species Coturnix gallica, all the medium-sized forms in the species Coturnix longipes, and all the Plio-Pleistocene forms in the recent species Coturnix coturnix. However, U. Göhlich has revised all these forms and shown that the genera Palaeortyx and Palaeoperdix are very distinct from each other and from Coturnix. In the recent genus Alectoris, Mlíkovsky places (pp. 157) two recent genera, Tetraogallus and Ammoperdix, the extinct genera Palaeocryptonyx, Pliogallus, Plioperdix, and Chauvireria, some of the species described in the genus Palaeoperdix, and many fossil species of the recent genus Francolinus. I have before me the type-material of Palaeocryptonyx and a sample of Chauvireria and can assert that these two genera are different from each other, and different from Alectoris.
Then, Mlíkovsky assigns a great number of Late Neogene birds to modern species. For example, among the grouse, he puts all the species (Tetrao praeurogallus, T. conjugens, T. macropus, T. rhodopensis, T. partium, Bonasa praebonasia, Lagopus balcanicus, L. atavus) described from the Early Pliocene (MN 14) to the Early Middle Pleistocene (MNQ 22) in the recent species Tetrao urogallus, T. tetrix, B. bonasia, and L. lagopus (pp. 166). Yet it has previously been demonstrated that each of those taxa is characterized by dimensions or proportions that are different from those of recent forms. By synonymizing all of the fossil taxa, Mlíkovsky suppresses all the information about the phylogenetic lineages that have led to the existing taxa.
From page 227 on, Passeriformes attributed by previous authors to recent species, or to recent genera but with indeterminate species, such as Erithacus spp., Luscinia spp., Oenanthe spp., Turdus spp., are considered by Mlíkovsky to be validly identified, and all the extinct species of Passeriformes are either synonymized with recent species or placed in a Family incertae sedis. In the family Corvidae, only one extinct species is retained— Miocorvus larteti, from the Middle Miocene—but even then Mlíkovsky says that “the taxonomic position of this species requires re-examination” (pp. 231). For all the large Plio-Pleistocene corvids, Mlíkovsky has tried to force the complex and multitudinous reality of the fossil into three mental categories corresponding to three recent species. Thus, he synonymizes Corvus fossilis, C. pliocaenus janossyi, and C. antecorax with C. corax; C. pliocaenus, C. praecorax, C. betfianus, and C. simionescui with C. corone; and Pica pica major (in part) and C. moravicus (described by Mlíkovsky himself) with C. monedula, while C. hungaricus is placed in the Aves incertae sedis. However, it is impossible to place in synonymy all those species, which have different sizes and display different proportions. By doing this, Mlíkovsky obfuscates the evolutionary sequences in the fossil record. Among the noncorvids there is a single exception, Loxia patevi Boev 1999, that is considered to be a valid extinct species of small passerine rather than Family incertae sedis (pp. 246). The other species described by Boev—Coccothraustes balcanicus, C. simeonovi, and Regulus balcanicus—have not benefited from the same leniency.
When several species, contemporaneous or spread out in time, are distinguished only by size, Mlíkovsky synonymizes them, even though one species may replace another biostratigraphically. When two species have a size difference on the order of 10%, he puts them into the same size class (pp. 90), but sometimes the size differences are much larger. For example, among the different species of the genus Idiornis, Mlíkovsky synonymizes I. gallicus, I. cursor, and I. gaillardi (pp. 180). Yet I. gallicus is on average 15% larger than I. cursor, and 80% larger than I. gaillardi. In I. gallicus, the total length of the tarsometatarsus is 108 mm, whereas in I. gaillardi the total length is 51 mm (Mourer-Chauviré 1983). On the other hand, in the genus Laricola, which Mlíkovsky created for the species Larus elegans Milne-Edwards and Larus totanoides Milne-Edwards, the two species are conserved (pp. 136), though Larus totanoides is only 5.6% larger than Larus elegans.
Although Mlíkovsky's action in this work was directed toward greatly reducing the number of taxa, at the same time he has created taxa that are unfounded. For example, the genus Anatalavis was created for a humerus from the Cretaceous or Paleocene of New Jersey, with the type-species Anatalavis rex. Olson (1999) described another species, A. oxfordi, from the Early Eocene of England, and indicated that the humerus of A. oxfordi “is similar in proportions and has the same distinctive curvature and robustness as that of A. rex…which it matches in all details except size.” Olson further specified, in a figure caption, that the slight differences between the photographs of the two humeri of A. oxfordi and A. rex “are mainly due to slightly different rotations of the specimens.” In spite of these indications, Mlíkovsky (pp. 107) uses those differences to establish a new genus, Nettapterornis.
In conclusion, this book is very dangerous. It has the appearance of a scholarly work, because of all the information given. Young avian palaeontologists taking their first steps in the discipline and neo-ornithologists interested in the evolution of birds will be tempted to consult it, but they will not be able to tell whether attribution of a fossil taxon to a given family or order comes from a general consensus, admitted by the international community, or from a purely arbitrary decision of the author.
This review gives a certain number of examples of disputable systematic positions and erroneous or arbitrary synonymizations, but it is by no means exhaustive, and I did not attempt to check everything written in this book. I hope, however, that my critique will convince readers to be very cautious when using it.