Pleistocene effects on North American songbird evolution

Recent studies have used comparisons of mitochondrial DNA (mtDNA) sequence divergence among populations and species to test existing hypotheses about avian evolution during the Pleistocene epoch. In 1998, Avise and Walker con cluded that the Pleistocene was an important time for avian evolution, incl uding the initiation of phylogeographic separations and the completion of s peciation events that began in the Pliocene. The study implied that these c onclusions conflicted with the study, in 1997, by Klicka and Zink, which co ncluded that most species pairs previously thought to have originated in th e past 250 000 years ct:ere much older. The two studies are complementary i n the sense that Avise and Walker dealt primarily with phylogeographic (int raspecific) separations. Furthermore, Klicka and Zink concentrated on the i nception of divergences whereas Avise and Walker focused on the timing of t he completion of speciation. To accomplish this, Avise and Walker analysed 'phylogroups', geographically coherent subsets of biological species in whi ch mtDNA haplotypes exhibit reciprocal monophyly. The study used the averag e interphylogroup mtDNA distance (0.027), calibrated at 2% per million year s, to conclude that speciation required on average one million years to com plete, Hence, speciation events begun in the Late Pliocene would have been completed in the mid- to late Pleistocene. Although we appreciate the exten ded nature of the speciation process and Avise and Walker's insightful atte mpt to estimate its duration, we conclude that their value was an overestim ate by a factor of two. In particular we question whether phylogroups can b e used in the novel evolutionary role that Avise and Walker envisioned, bec ause of the vagaries of taxonomic practices and lack of consensus regarding species concepts. To extend their analysis of intraspecific, phylogeograph ic separations, we compiled previously analysed and newly available data fo r divergence times for North American songbird (order Passeriformes) phylog roups. More than 80% were initiated at least one million years ago, which i s inconsistent with the late Pleistocene origins model previously rejected by Klicka and Zink. Although some divergence events can be traced to the la te Pleistocene, the significance of the distribution must be judged with re ference to a null model. Whether the Pleistocene was a profound time for av ian phylogeographic differentiation is at present unknown.