Format: ORAL

Herv Sauquet1,2,3, Andrea M. Lpez-Martnez4, Maria von Balthazar5, Marion Chartier5, Julian Herting1, Santiago Ramrez-Barahona4, Susana Magalln4, Jrg Schnenberger5

1 Botanic Gardens of Sydney, Mount Annan, Australia 2 University of New South Wales, Sydney, Australia 3 University of Sydney, Sydney, Australia 4 Universidad Nacional Autónoma de México, Ciudad de México, Mexico 5 University of Vienna, Vienna, Austria

The eFLOWER project started thirteen years ago as a long-term initiative to address key questions on the early evolution and diversification of angiosperm flowers. A central part of eFLOWER has been the development and continuous curation and expansion of a collaborative database named PROTEUS allowing to record floral traits at the species level while documenting the explicit source of each data point. Here we review the diverse contributions that eFLOWER has brought so far to our understanding of floral evolution, reflect upon the questions and challenges encountered along the way, and outline future directions. An important milestone was reached seven years ago with the publication of a new model for the ancestral flower of all angiosperms, based on ancestral state reconstructions using a dataset of 792 species sampled from most families. Despite uncertainty remaining for some traits, this new model proved effective in questioning long-term assumptions and opening new research opportunities on floral evolution. Interestingly, this model appears to be robust to new analyses of a recently published, expanded dataset of 1201 species sampled from all families, but some results continue to be highly sensitive to methodological approaches and models of character evolution. Another significant milestone has been the development and evaluation of angiosperm-wide phyloscan approaches to assess the phylogenetic position of fossil flowers using these datasets of floral traits. While these methods appear to provide a powerful new tool to systematically assess fossil flowers, they have also revealed high uncertainty in the placement of many taxa, reflecting the limits of the number of preserved, standardised characters that may be recorded from fossil flowers. Lastly, we present briefly some recent new avenues linked with this project, including new ideas on the origin of pollination and evidence on early angiosperm diversification from morphospace analyses of floral disparity.