Aim Various data sets and methods of analysis were combined to produce the first comprehensive molecular phylogeny of the genus Tuber and to analyse its biogeography.
Location Europe, North Africa, China, Asia, North America.
Methods Phylogenetic relationships among Tuber species were reconstructed based on a data set of internal-transcribed spacer (ITS) sequences and various phylogenetic inference methods, specifically maximum parsimony, Bayesian analysis and neighbour joining. Tajima’s relative rate test showed that Tuber 18S rRNA, 5.8S rRNA, 5.8S-ITS2 rRNA and β-tubulin sequences evolved in a clock-like manner. These genes, combined or not, were employed for molecular clock estimates after construction of linearized trees using mega 3.1. We reconstructed ancestral areas in the Northern Hemisphere by means of a dispersal–vicariance analysis (diva 1.1) based on current distribution patterns of the genus Tuber determined from the literature.
Results The resulting molecular phylogeny divided the genus Tuber into five distinct clades, in agreement with our previously published studies. The Puberulum, Melanosporum and Rufum groups were diversified in terms of species and geographical distribution. In contrast, the Aestivum and Excavatum groups were less diversified and were located only in Europe or North Africa. Using a global molecular clock analysis, we estimated the divergence times for the origin of the genus and for the origin of several groups. diva inferred nine dispersal events and suggested that the ancestor of Tuber was originally present in Europe or was widespread in Eurasia. Equally optimal distributions were obtained for several nodes, suggesting different possible biogeographical patterns.
Main conclusions Our analyses identified several discrepancies with the classical taxonomy of the genus, and we propose a new phylogenetic classification. According to molecular clocks, the radiation of the genus Tuber could have started between 271 and 140 Ma. Used in combination with the results obtained from time divergence estimates, this allows us to propose two equally probable scenarios of intra- and inter-continental diversification of the genus according to the geographic distribution of the most recent common ancestor in Europe or Eurasia. The biogeographical patterns imply intra-continental dispersal events between Europe and Asia and inter-continental dispersal events between North America and Europe or Asia, which are compatible with land connections during the Tertiary.