Format: ORAL

Douglas E. Soltis

Florida Museum of Natural History, University of Florida, Gainesville, FL, USA

The first four iterations of the APG classification were largely based on sequence data from the plastid genome. Over the past decade, the use of numerous nuclear loci across many taxa has become increasingly common in phylogenetic analyses of plants. In many cases, nuclear and plastid-based topologies for angiosperms agree, providing increased support for inferred relationships. However, in other cases, comparison of nuclear and plastid-based topologies has revealed surprising instances of incongruence at levels much deeper than those likely resulting from recent hybridization or incomplete lineage sorting over recent timescales. These examples of conflict provide important insights into evolution, showing that reticulation has been common and widespread throughout angiosperm evolutionary history. But how best to reflect these examples of deep incongruence in classification, especially in the next APG classification?Various approaches deserve consideration as we attempt to construct classifications that are maximally informative about phylogenetic history. When nuclear and plastid phylogenies conflict, the classification could alternatively follow: (1) the nuclear topology, with symbols indicating where the plastid genome provides an alternative topology; or (2) both the nuclear and plastid topologies by placing a taxon in alternative clades (as allowed under the PhyloCode). Each of these approaches presents challenges. For example, what percentage of nuclear genes would be required for a particular placement to be the favored placement in APG? And what navigation tools might be needed to alert users that a taxon is placed in two alternative groups? A system that provides maximum information and flexibility should be employed.