The two recent theories of flowering plant evolutionary origins—the Anthophyte and Neo-Pseudanthial theories—are based on phylogenies (from morphological data) that show Gnetales as extant sister to angiosperms or as paraphyletic to angiosperms. Gnetales figured prominently in homology assessments and evolutionary scenarios of these theories. Several recent studies, including ours, provide strong evidence that extant gymnosperms are monophyletic, so Gnetales are most closely related to other gymnosperms, not to flowering plants. This removes the basis for both recent theories, leaving earlier theories that relate flowering plants to fossil “seed-fern” gymnosperms as the only active contenders. Our data from the homeotic gene Floricaula/LEAFY imply that the lineage leading to flowering plants originally had two copies of this gene, but that one copy was lost in flowering plants, which suggests a new theory: that developmental control of flower organization derives more from systems active in the male reproductive structures of the gymnosperm ancestor, rather than from the female, with ovules being ectopic in the original flower. This theory was based entirely on data from living plants, but the fossil group Corystospermales has both male and female structures that fit the theory. Corystosperms could include ancestors of the flowering plants.

Communicating Editor: Paul Wilson