Fossil calibration of Magnoliidae: thorough background research significantly improves reliability of molecular age estimates

Author(s)
Julien Massoni, Maria von Balthazar, Laetitia Carrive, Thomas Couvreur, Jürg Schönenberger, Yannick Städler, Hervé Sauquet
Abstract

In order to obtain divergence times within a group, a diversity of approaches is conceivable. Recent research has shown that good fossil calibration is critical when using molecular dating methods. The use of fossils to provide minimum age constraints requires a thorough review of their phylogenetic placements and of their geological ages. Here we focus on the estimation of divergence times within Magnoliidae (Canellales, Laurales, Magnoliales, Piperales). This clade of angiosperms, comprising 10,000 species, presents a rich fossil record dating back to the Early Cretaceous. A wide range of estimates of divergence times for deep nodes of the group has previously been obtained in large-scale studies comprising a maximum of 36 composite taxa of Magnoliidae and six internal calibration points. Here we propose a new timetree of Magnoliidae based on two relaxed-clock analyses (PL, UCLN) of a molecular data set of 12 markers from the three genomes and 221 exemplar species representing 84% of the genera recognized in the clade. To calibrate our analyses, we have reviewed the age and the phylogenetic placement of 10 fossils of Magnoliidae, leading to a revised calibration scheme of 8 effective minimum age constraints. In addition, we have built a morphological data set of 75 floral characters recorded for the same species as in the molecular data set as well as eight fossil species of Magnoliidae. Together, these two data sets (molecular and morphological) allowed us to evaluate the position of these fossils in a new phylogenetic context using total evidence and molecular backbone approaches. In addition to traditional molecular dating approaches using prior calibration schemes (“node-dating”), our combined data set give us the opportunity to conduct molecular dating incorporating both extant and extinct species as terminal taxa (“total-evidence dating”). Our results suggest that both the thorough review of fossil calibrations and the re-analysis of their relationships in a modern phylogenetic framework (requiring the assembly of morphological data sets) can significantly improve the accuracy of molecular divergence times estimated with relaxed-clock methods.

Organisation(s)
External organisation(s)
Université Paris XI - Paris-Sud, Institut de recherche pour le développement
Publication date
2013
Austrian Fields of Science 2012
106042 Systematic botany, 106012 Evolutionary research, 106008 Botany
Portal url
https://ucrisportal.univie.ac.at/en/publications/2172501d-1952-4e0b-a2f2-6ef7a593233b