Taxon Sampling and the Phylogenetic Position of Passeriformes: Evidence from 916 Avian Cytochrome b Sequences

Abstract

Attempts to resolve relationships among major groups of birds from DNA sequences have been plagued by the problem of apparent long branches connected by short internodes (Sibley and Ahlquist, 1990) and the difficulties of sampling a large number of taxa in one study. A recent study of relationships among major avian lineages based on complete mitochondrial DNA sequences (Mindell et al., 1997, 1999) produced a striking result of a basal position for passerine birds (Aves: Passeriformes), placing them as sister to all other birds. This arrangement is novel and differs from Sibley and Ahlquist's (1990) tree based on DNA-DNA hybridization data and from virtually all previous classification schemes for birds (Fig. 1). Most classifications of birds place Passeriformes well within other neognath lineages (e.g., Morony et al., 1975; Howard and Moore, 1980; del Hoyo et al., 1992; Groth and Barrowclough, 1999) and place paleognaths (ratites and tinamous) as the sister to neognaths (all other birds). One potential criticism of Mindell et al.'s (1997, 1999) studies is that taxon sampling is very sparse (five and seven avian species, respectively). A study with greater taxonomic coverage (21 avian species) based on only mitochondrial cytochrome b (cyt b) DNA sequences (Harlid et al., 1998) still resulted in a sister relationship between passerines and all other birds. Nonetheless, all of these DNA sequence studies include only a very small frac-