View article

METHODS

Lynch's method requires two types of data: explicit phylogenetic hypotheses and distribu-tional information for monophyletic groups of species. We analyzed phylogenies of 13 lineages of birds that satisfied these criteria (table 1); many published phylogenetic analyses were excluded because they failed to examine all members of a clade. As a further test of Lynch's method, we compiled a list of 20 presumed sister species of North American passerine birds (table 2) and estimated speciation mode in these pairs. Fol-lowing Lynch, we also considered only conti-nental taxa. The taxonomic and latitudinal spec-