PREMISE OF THE STUDY:  When fossils are sparse in morphologically divergent lineages, phylogenetic analyses based on morphology may support hypotheses of relationships incongruent with those supported by gene sequence data. Integration of morphological and sequence data from conservative gene regions may reconcile such situations by circumscribing the search space of combined analyses.

METHODS:  We revisited the phylogeny of Polytrichopsida, a highly divergent group of mosses, in parsimony analyses. We supplemented the existing morphological matrix with sequence data of four genes (nt 18S, cp rbcL & rps4, mt nad5) commonly used in mosses; we included both fossils and extant terminals that lack sequence data.

KEY RESULTS:  We recovered patterns of relationships largely congruent with those supported by analyses of sequence data from a higher number of terminals. Analyses including all terminals recovered one optimal tree with Alophosia and the fossil Meantoinea forming a clade sister to the rest of the ingroup. Outgroup sampling experiments produced congruent topologies for most ingroup relationships.

CONCLUSIONS:  The potential of total-evidence analyses sampling conservative genes to alleviate conflict with morphology-based hypotheses should be tested in groups with denser fossil record. Although bryophyte fossils have been deemed fragmentary, it was recently shown that integrating fossils in total-evidence analyses may not reduce phylogenetic resolution. Therefore, total-evidence analyses incorporating fossils bear promise for clarifying bryophyte phylogenetic relationships. In Polytrichopsida, a group with a long evolutionary history, we cannot assume, a priori, that currently known fossils, which date back to only c. 140 Ma ago, represent the oldest lineages.

TNT-file with 11 continuos additive characters, followed by a combined matrix of 89 non-additive chars. of morphology; 83 of 18S, 254 rbcL, 158 rps4 & 40 of nad5