Integration of Paleobotanical and Neobotanical Data in the Assessment of Phytogeographic History of Holarctic Angiosperm Clades

Attempts to reconstruct phytogeographic history based exclusively on either modern or fossil distribution patterns may give misleading results. Local extinction within a widespread clade can undermine phylogeographic hypotheses based on modern‐day distribution patterns; e.g., within the Aceroideae, Dipteronia is the sister group to Acer and is presently known only from Asia; the discovery of Dipteronia in the North American Tertiary requires a rethinking of hypotheses based solely on the modern distribution of the two genera. Conversely, the fossil record alone cannot provide a well‐resolved phytogeographic history because the availability of well‐preserved fossils for a particular time interval differs among regions. The most informative method for deducing the phytogeographic history of a given clade is to conduct phylogenetic analyses that include both fossil and extant representatives to deduce the sequence of dispersal and/or vicariance events indicated by the most parsimonious trees. The choice of taxa for such investigations should be at least partially based on the richness and quality of paleontological data. The Ulmaceae sensu stricto are cited as an example to illustrate the effect of including versus excluding an extinct genus from morphological phylogenetic analyses. Cedrelospermum (Tertiary of southern Mexico through midlatitude North America and Europe) is resolved as a link, both phylogenetically and geographically, between the Central and South American taxa Ampelocera and Phyllostylon and the genera of North America, Europe, and Asia. Inclusion of the extinct taxon in the data matrix along with the extant taxa significantly alters the resulting phylogeographic model.