Phylogenetic biogeography of the leafy liverwort Herbertus (Jungermanniales, Herbertaceae) based on nuclear and chloroplast DNA sequence data: correlation between genetic variation and geographical distribution

Abstract Aim  The cosmopolitan genus Herbertus is notorious for having a difficult taxonomy and for the fact that there is limited knowledge of species ranges and relationships. Topologies generated from variable molecular markers are used to discuss biogeographical patterns in Herbertus and to compare them with the geological history of continents and outcomes reported for other land plants. Location  Africa, Asia, Azores, Europe, southern South America, northern South America, North America, New Zealand. Methods  Phylogenetic analyses of nuclear ribosomal internal transcribed spacer and chloroplast (cp) trn L– trn F sequences of 66 accessions of Herbertus and the outgroup species Triandrophyllum subtrifidum and Mastigophora diclados were used to investigate biogeographical patterns in Herbertus . Areas of putative endemism were defined based on the distribution of species included in the analyses. Maximum parsimony analyses were undertaken to reconstruct ancestral areas and intraspecies migration routes. Results  The analyses reveal species‐level cladograms with a correlation between genetic variation and the geographical distribution of the related accessions. The southern South American Herbertus runcinatus is sister to the remainder of the genus, which is split into two main clades. One contains the Neotropical–African Herbertus juniperoideus and the New Zealand/Tasmanian Herbertus oldfieldianus . An African accession of H. juniperoideus is nested within Neotropical accessions. The second main clade includes species that inhabit Asia, the Holarctic, Africa, and northern South America. Maximum parsimony analyses indicate that this clade arose in Asia. Herbertus sendtneri originated in Asia and subsequently colonized the Holarctic and northern South America. An Asian origin and colonization into Africa is indicated for H. dicranus . Main conclusions  The current distribution of Herbertus cannot be explained by Gondwanan vicariance. A more feasible explanation of the range is a combination of short‐distance dispersal, rare long‐distance dispersal events (especially into regions that faced floral displacements as a result of climatic changes) extinction, recolonization, and diversification. The African Herbertus flora is a mixture of Asian and Neotropical elements. Southern South America harbours an isolated species. The molecular data indicate partial decoupling of molecular and morphological variation in Herbertus . Biogeographical patterns in Herbertus are not dissimilar to those of other groups of bryophytes, but elucidation of the geographical ranges requires a molecular approach. Some patterns could be the result of maintenance of Herbertus in the inner Tropics during glacial maxima, and dispersal into temperate regions in warm phases.