The Howard and Moore Complete Checklist of the Birds of the World, 3rd edition.—Edward C. Dickinson, Ed. 2003. Princeton University Press, Princeton, New Jersey. 1,039 pp. ISBN 0-691-11701-2. Cloth, $69.50.—In the past half-century, ornithologists have revolutionized taxonomy and systematics, created several new species concepts, and recognized dozens of new species-level avian taxa. Yet the ornithological community has been without a scholarly and functional standard classification and checklist of the world's birds since the 15-volume Peters checklist was finished in 1974 (with half the volumes now more than 45 years old). As global communication rises exponentially, scientists and globetrotting birdwatchers have been in dire need of a standard classification to serve as the lingua franca of ornithology.

With publication of The Howard and Moore Complete Checklist of the Birds of the World, edited by E. C. Dickinson, we now have a thorough and useful book that gives a modern classification of all the world's birds, down to the subspecific level, in a single volume. Checklists produced in the interlude between Peters and Dickinson fell short for a number of reasons, including out-of-date classifications, lack of subspecific treatment, overly novel classification schemes, and partial to complete lack of references. Justifications for taxonomic treatments have been all but absent in those volumes, especially at levels other than species. Dickinson does not suffer from those drawbacks and should serve as a standard reference for bird classification for the next little while. This volume arrived at an opportune time for our work at the Florida Museum of Natural History, as we are using a recent move into an enlarged space to install our collections in a more modern sequence. I therefore had a golden opportunity to use and review this volume.

The introduction outlines three objectives. The first is to provide a comprehensive list, which necessitated including all newly described species, whether recognized or in synonymy. Regional consultants (E. C. Dickinson for Asia, D. Pearson for Africa, J. V. Remsen, Jr., for the Americas, K. Roselaar for the Palaearctic, and R. Schodde for Australasia) helped to ensure completeness. Those subregional editors have worked within or closely with continent-wide committees on avian classification (e.g. Remsen in the AOU Committee on Classification and Nomenclature and the South American Classification Committee). The second objective is to present a conservative list. Using the Peters Checklist as a foundation, the consultants accepted changes only when there had been “persuasive published reasons.” The third objective is a high standard of nomenclatural accuracy. Curiously, stability was not included as an objective, though the three stated objectives all work toward that goal. The introduction also gives explanations of list sequence, species concepts, taxon recognition, scientific names, English names (thankfully not a main focus of revisions), list of references, and cut-off date for incorporation of new material (31 December 2000).

A separate introductory chapter, written by J. Cracraft, F. K. Barker, and A. Cibois from the American Museum of Natural History, gives a quick overview of higher-level phylogenetics (above the rank of family) and the rationale for some of the novel treatments in this volume, especially the sequence of families and the non-use of taxa above family (see below).

A list of families follows. The list sequence of families generally follows some of the more recent advancements in avian higher-level systematics, while retaining some similarity to more classical schemes. Some of the changes to traditional sequences have become widely accepted in recent years (e.g. beginning the sequence within the Neoaves with Galliformes and Anseriformes), whereas others may be more difficult to get used to, and may or may not gain future acceptance. Some of the changes seem premature and out of step with the conservative objectives of the book. In their brief synopses of ordinal- and family-level systematics, Cracraft et al. provide details on the placement of more controversial taxa (e.g. Catharidae, Phoenicopteridae) but usually do not provide justifications for sequences within orders. For example, the first family listed for the traditional Falconiformes is the Falconidae (usually placed last within that order), and Bucconidae and Galbulidae are listed after other families in the Piciformes (usually placed before). Justification for the sequence in the Falconiformes is not given, and because the order does not have a clear close relative to serve as an outgroup, establishing the basal branching pattern is highly speculative. A more conservative approach would have been to keep the traditional sequence. In their explanation of the taxonomy of the Pici (Piciformes and relatives), Cracraft et al. state that there are two clear taxa (toucan plus barbets and honeyguides plus picids) and that Galbulidae may be more closely related to the Coraciiformes. The Bucconidae are not mentioned in their synopsis. A conservative sequence would thus be similar to traditional sequences (Galbulidae, Bucconidae, Ramphastidae, Indicatoridae, Picidae), but the family sequence they present (Ramphastidae, Indicatoridae, Picidae, Galbulidae, Bucconidae) appears to lack support and has little hope of long-term stability. A more egregious veering away from the traditional sequence makes little sense: the New World suboscine passerines consist of two related groups of families, the funariid group (roughly Furnariidae-Dendrocolaptidae-Thamnophilidae-Formicariidae-Rhinocryptidae-Conopophagidae) and the tyrannid group (Tyrannidae-Pipridae-Cotingidae). With a two-taxon statement, it does not matter which group is first; thus, the traditional sequence with the furnariid group first could have been preserved. Instead, Cracraft et al. begin with the tyrannid group of families.

The main list follows. The author and date are given for each genus, species, and subspecies. I loudly applaud the inclusion of full citations for all taxa described since the Peters Checklist; these are given in the list of references. Each genus includes its gender, which should help describers of new taxa. Each species has an English name, and, if monotypic, its distribution is given. If a species has multiple subspecies, distributions are described briefly. All stated distributions are highly abbreviated, and almost always are given in political units. For large units (e.g. Mexico, California) the distributions are excessively abbreviated; for example, the distributions of four subspecies of Melospiza melodia are “coastal c[entral] California,” but there was ample space to put, for example, “N San Francisco Bay.” Nevertheless, abbreviation pays off in having everything fit into a tight volume. Between 25 and 45 terminal taxa (sub-species and monotypic species) are included per page. The copious comments are footnoted (up to 22 per page), which include standard author and date citations. Full citations (2,739 of them) are in the list of references at the end of the list. Taxonomic names are completely indexed, so that genus and species are given for each sub-specific name.

I found the decision not to include taxa above family level a major disappointment in the checklist. Thus, you will not find mention of such well-established higher taxa as Passeriformes, Anseriformes, Galliformes, Falconiformes, Strigiformes, and Procellariiformes. The families that compose these and most other orders are well known, though the relations among orders are still far from clear. Cracraft et al. explain in the introductory chapter on the subject that higher-level relationships among birds “are still clouded with uncertainties.” However, what taxonomic level is not clouded by such uncertainties? Though Cracraft et al. spell out many of the better-documented higher-level relationships in their introductory chapter, the checklist should have used an ordinal-level classification.

With such a large, data-rich volume, it is not hard to find minor problems and quibbles with the text; I will mention just a few. Distribution errors are especially numerous. For example, the distributions for the monarchids Neolalage banksiana and Clytorhynchus pachycephaloides grisescens are both given as the Banks Islands, within Vanuatu, when in fact both taxa are distributed throughout the Vanuatu archipelago. In the distribution of the fantail Rhipidura rufifrons, the large islands of Chousieul and Santa Ysabel (= Isabel) in the Solomon Islands are missing; they are inhabited by R. r. commoda. Dickinson follows the ridiculous lumping of the flightless rail Nesoclopeus woodfordi of the Solomon Islands into the equally flightless N. poecilopterus of Fiji, though thousands of kilometers and several very deep ocean trenches separate the two taxa. In splitting up what was the world's most geographically variable species (Pachycephala pectoralis), one of the eight resulting species (citing personal communication from R. Schodde, but not keeping to the goal of “persuasive published reasons”) is given the English name New Caledonian Whistler, even though it also occurs in Vanuatu. Dickinson neglects to cite T. A. Parker (1982, Wilson Bulletin 94:484) for discovering that the recently described antbird Percnostola macrolopha Berlioz, 1966, was actually the female of P. lophotes. Dickinson does not provide justification for retaining North American parids in Parus instead of placing them in Poecile (chickadees) or Baeolophus (titmice). In general, Dickinson closely follows the classification presented in the seventh edition of the AOU Check-list of North American Birds, which recognized Poecile and Baeolophus.

The sheer volume of molecular systematics papers nowadays is rocking the boat of avian classification and will, no doubt, cast asunder much of current taxonomy and classification in the future. Workable classifications and checklists will have to be able to adapt to new arrangements as published evidence becomes available, but also will have to retain some stability to be useful. The only way to do this is to be thorough, be conservative in changes, and provide justification and citations for taxonomic treatments. Dickinson has set a high standard for how checklists should be written. I hope to see future editions in a similar style.