[Description of methods used for collection/generation of data] We built a genus-level phylogeny using the workflow proposed by Roquet et al. (2013). We downloaded from Genbank three conserved chloroplastic regions (rbcL, matK and ndhF) plus the ITS region for a subset of families, which we aligned separately by taxonomic clustering. [Methods for processing the data] We aligned all coding sequence clusters with MACSE (Ranwez et al. 2011) and non-coding ones with MAFFT (Katoh and Standlye 2013), and trimmed all alignments with TrimAl (Capella-Gutiérrez et al. 2009). We concatenated all alignments to obtain a supermatrix. We then conducted maximum-likelihood (ML) phylogenetic inference analyses with RAxML (Stamatakis 2014), applying the most appropriate partitioning scheme and substitution model obtained with PartitionFinder (Lanfear et al. 2012) and a supertree constraint at the family-level obtained with the online software Phylomatic v.3 (tree R20120829). Specifically, we performed 100 independent tree searches and selected the best ML tree (the one with the highest probability). [Instrument- or software-specific information needed to interpret/reproduce the data] Any software for manipulation or analysis of phylogenies such as R packages ape or picante. [Standards and calibration information, if appropriate] The best ML tree was dated applying the penalized likelihood method in treePL (Smith and O'Meara,2012) and the following node calibrations: we fixed the node corresponding to the ancestor of eudicots at 125 Ma based on the earliest eudicot fossil (Hughes and McDougall 1990), and applied minimum age constraints to 15 nodes based on fossil information extracted from Smith and Beaulieu (2010) and Bell et al. (2010). 5. Environmental/experimental conditions:

Agencia Estatal de Investigación, Project VULVIMON (Reference: CGL2017-90040-R).

Peer reviewed