TL;DR: A phylogenetic analysis of a combined data set for 560 angiosperms and seven outgroups based on three genes, 18S rDNA, rbcL, and atpB representing a total of 4733 bp is presented, resulting in the most highly resolved and strongly supported topology yet obtained for angiosPerms.
TL;DR: Evidence from morphology and molecular sequence data indicates that extant Ranunculidae as well as “lower” Hamamelididae and several other families are basal in this group of dicotyledons, and cyclic floral architecture is probably basic for eudicots as a whole.
Abstract: Recent cladistic analyses of angiosperms based on both morphological and molecular sequence data recognize a major clade of dicotyledons defined by triaperturate or triaperturate-derived pollen (non-magnoliids/eudicots). Evidence from morphology, as well as the atpB and rbcL genes (cpDNA), indicates that extant Ranunculidae (e.g., Papaverales, Lardizabalaceae, Berberidaceae, Menispermaceae, Ranunculaceae) as well as “lower” Hamamelididae [e.g., Eupteleaceae (allied to Ranunculidae), Hamamelidaceae, Myrothamnaceae, Platanaceae, Trochodendraceae] and several other families (e.g., Gunneraceae, Nelumbonaceae, Proteaceae, Sabiaceae) are basal in this group. The earliest records of diagnostic eudicot pollen are of mid-late Barremian age (c. 126myr BP) and by around the latest Albian (c. 97 myr BP) several basal eudicot groups (e.g., Trochodendrales, Platanaceae, Buxaceae, and perhaps Circaeasteraceae, Myrothamnaceae, and Nelumbonaceae) are recognizable in the fossil record. Possible Hamamelidaceae and perhaps Proteaceae are present by the Turonian (c. 90 myr BP). Among basal eudicots, flowers are generally bisexual although unisexual flowers are also common. In some groups (e.g., Myrothamnaceae, Buxaceae, certain Berberidaceae), delimitation of the flower is not always clear and there is a more or less gradual transition between tepals and inflorescence bracts. Plasticity in floral form at this level of angiosperm evolution is predominantly encompassed by dimerous and trimerous cyclic floral organization and transitions from one to the other are common. Spiral floral phyllotaxis of numerous stamens and carpels is more or less restricted to the Ranunculaceae. The basic condition of the perianth in eudicots appears to lack differentiation into sepals and petals, and petals appear to have arisen independently numerous times from stamens. Based on the generality of its systematic distribution, cyclic floral architecture is probably basic for eudicots as a whole, and at this level of angiosperm evolution flowers with numerous, helically-arranged stamens and/or carpels (e.g., many Ranunculaceae) almost certainly reflect processes of secondary multiplication that have occurred independently many times.
TL;DR: Maximum likelihood phylogenetic analysis of a 79-gene, 97-taxon data set that included all available early-diverging eudicots and representative sampling of remaining angiosperm diversity largely agreed with previous estimates of early-Diverting eudicot relationships, but resolved Trochodendrales rather than Buxales as sister to Gunneridae, albeit with relatively weak bootstrap support, conflicting with what has been found for these three clades in most previous analyses.
TL;DR: Gynoecium and ovule structure was compared in representatives of the basal eudicots of Ranunculales, Proteales, and some families of the former «lower» hamamelids that have been moved to Saxifragales and incertae sedis to support the close relationship of Platanaceae and Proteaceae.
TL;DR: The character reconstructions indicate that the woody habit is ancestral, not only for the basal angiosperms, but also for the eudicots, and Ranunculales may not be ancestrally herbaceous, as long maintained.