A PHYLOGENETIC ANALYSIS OF BASAL ANSERIFORMES, THE FOSSIL PRESBYORNIS, AND THE INTERORDINAL RELATIONSHIPS OF WATERFOWL

A phylogenetic analysis of 123 morphological characters of basal water fowl (Aves: Anseriformes) and other selected avian orders confirmed th at the screamers (Anhimae. Anhimidae) are the sister-group of other wa terfowl (Anseres), and that the magpie goose (Anseranatidae: Anseranas semipalmata) is the sister group of other modern waterfowl exclusive of screamers (Anatidae sensu stricto). The analysis also supports the traditional hypothesis of the gallinaceous birds (Galliformes) as the sister group of the Anseriformes. Presbyornis, a fossil from the early Eocene of Wyoming and averred by Olson & Feduccia as showing that the Anseriformes were derived from shorebirds (Charadriiformes), was foun d to represent the sister group of the Anatidae, Associated hypotheses by Olson gr Feduccia concerning the implications of Presbyornis for t he phylogenetic relationships of flamingos (Phoenicopteriformes), the position of the Anhimidae within the waterfowl, relationships among mo dern Anatidae, and a plausible evolutionary scenario for waterfowl als o are rejected. Analyses revealed that cranial characters were critica l to the establishment of the Galliformes as the sister group of the A nseriformes; exclusion of the Anhimidae, especially in combination wit h Anseranas, also undermined the support for this inference. Placement of Presbyornis as the sister group of the Anatidae casts doubt on the role suggested by Feduccia of 'transitional shorebirds' in the origin of modern avian orders, and calls into question the concept of 'fossi l mosaics'. The phylogenetic hypothesis is used to reconstruct an evol utionary scenario for selected ecomorphological characters in the gall iform-anseriform transition; to predict the most parsimonious states o f these characters for Presbyornis, and to propose a phylogenetic clas sification of the higher-order taxa of waterfowl. This re-examination of Presbyornis also is used to exemplify the fundamental methodologica l shortcomings of the intuitive approach to the reconstruction of phyl ogenetic relationships. (C) 1997 The Linnean Society of London.