The phylogenetic relationships of the shags and cormorants: Can sequence data resolve a disagreement between behavior and morphology?

Taxonomic arrangements for the cormorants and shags (Phalacrocoracidae) had varied greatly until two quite similar arrangements, one based on behavior and the other on osteological characters, became the basis for current tho ught on the evolutionary relationships of these birds. The terms cormorant and shag, which had previously been haphazardly applied to members of the g roup, became the vernacular terms for the two major subdivisions within thi s family. The two taxonomies differ in places, however, with the behavioral . taxonomy placing several species within the shags and the osteological ta xonomy and phylogeny grouping those species (as the marine cormorants) and placing them within the cormorants. In an attempt to resolve the difference s in the relationships hypothesized by behavior and morphology, we sequence d three mitochondrial genes (12S, ATPase 6, and ATPase 8). Initial equally weighted parsimony trees differed slightly from our two weighted parsimony trees, one of which was also our maximum-likelihood tree. Many of the branc hes within our trees were well supported, but some sections of the phylogen y proved difficult to resolve with confidence. Our sequence trees differ su bstantially from the morphological phylogeny and show that neither the shag s nor the cormorants are monophyletic, but form an intermingled group. Some of the groups supported by both the behavioral and the morphological taxon omies (e.g, the cliff shags, Stictocarbo) appear to be polyphyletic. Conver sely, the monophyly of the blue-eyed shags, a traditional group that the os teological analysis had found to be paraphyletic, was supported by the sequ ence data. Until more taxa are sampled and a fully robust phylogeny is obta ined, a conservative approach accepting a single genus, Phalacrocorax, for the shags and cormorants is recommended. (C) 2000 Academic Press.