Background:

Echeveria and Pachyphytum are two closely related Neotropical genera in the Crassulaceae. Several species in Echeveria possess characters cited as diagnostic for Pachyphytum such as a clearly defined stem, a nectary scale on the inner face of petals and as inflorescence a scorpioid cyme or cincinnus. Pachyphytum has been identified as monophyletic while Echeveria as polyphyletic in previous molecular phylogenetic analysess.

Hypothesis:

The objective of this paper is to identify the phylogenetic position of a rare species with restricted distribution in Echeveria, E. heterosepala that possesses the diagnostic characters of Pachyphytum to better understand the generic limits between these two genera. We expect this species to be closely related to Pachyphytum.

Methods:

Bayesian inference and Maximum Likelihood analyses were carried out using 47 taxa, including as ingroup, species of Echeveria, Graptopetalum, Lenophyllum, Pachyphytum, Sedum, Thompsonella and Villadia and as outgroup, species in Dudleya. Analyses were conducted based on plastid (rpl16, trnL-F) and nuclear (ETS, ITS) markers. Ancestral character reconstruction was carried out under a parsimony criterion based on the molecular trees retrieved by the phylogenetic analyses. Four morphological characters were considered: defined stem, type of inflorescence, nectary scale in petals and position of sepals.

Results:

Accessions of E. heterosepala were retrieved in a well-supported clade group comprising exclusively Echeveria species. Therefore this species belongs to Echeveria although possessing characters similar to Pachyphytum and moreover it was not identified closely related to this genus. None of the traits considered have taxonomic importance. The node at the Pachyphytum clade identified unambiguous character states such as stem present, straight sepals, nectary scale on petals, however these character states were identified in the rest of the clades as well. Remarkably, the monophyly of Pachyphytum was corroborated, while Echeveria remains poorly understood.